From 0ebcf2e7cd43fa685952164884e16e732535cc4f Mon Sep 17 00:00:00 2001 From: jhudsl-robot Date: Tue, 11 Apr 2023 22:43:01 +0000 Subject: [PATCH] Render bookdown --- docs/01-intro.md | 130 ++++----- docs/02-intro.md | 67 +++++ docs/03-learning-from-failure.md | 13 + docs/04-case-study-genomic.md | 55 ++++ docs/404.html | 140 ++-------- ...Introduction-to-Reproducible-Research.docx | Bin 0 -> 180657 bytes docs/References.md | 2 +- docs/about-the-authors.html | 144 ++-------- ...-study-developing-a-genomic-predictor.html | 254 ++++++++++++++++++ docs/from-x-to-computational-x.html | 254 ++++++++++++++++++ docs/index.html | 148 ++-------- docs/index.md | 8 +- docs/introduction.html | 251 ++++------------- docs/learning-from-failure.html | 220 +++++++++++++++ docs/reference-keys.txt | 43 ++- docs/references.html | 161 ++--------- ...gR8P6iDX4DfFD65W_gdQ_gcc4fbee202_0_141.png | Bin 17318 -> 0 bytes .../figure-html/unnamed-chunk-4-1.png | Bin 19921 -> 0 bytes ...zTzFH4PHWzm9cuXbKpzlbSd1TvXEy60Sste0_p.png | Bin 0 -> 158114 bytes docs/search_index.json | 2 +- 20 files changed, 1072 insertions(+), 820 deletions(-) create mode 100644 docs/02-intro.md create mode 100644 docs/03-learning-from-failure.md create mode 100644 docs/04-case-study-genomic.md create mode 100644 docs/Introduction-to-Reproducible-Research.docx create mode 100644 docs/case-study-developing-a-genomic-predictor.html create mode 100644 docs/from-x-to-computational-x.html create mode 100644 docs/learning-from-failure.html delete mode 100644 docs/resources/images/02-chapter_of_course_files/figure-html/1YmwKdIy9BeQ3EShgZhvtb3MgR8P6iDX4DfFD65W_gdQ_gcc4fbee202_0_141.png delete mode 100644 docs/resources/images/02-chapter_of_course_files/figure-html/unnamed-chunk-4-1.png create mode 100644 docs/resources/images/02-intro_files/figure-html/1-pVlxwkzTzFH4PHWzm9cuXbKpzlbSd1TvXEy60Sste0_p.png diff --git a/docs/01-intro.md b/docs/01-intro.md index e5d42ea..21ef3c3 100644 --- a/docs/01-intro.md +++ b/docs/01-intro.md @@ -4,87 +4,53 @@ # Introduction +This chapter will be about the concept of reproducible research and we will cover some basic concepts and ideas that are related to reproducible research. +Before we get to *reproducibility*, we need to cover a little background with respect to how science works (even if you're not a scientist, this is important). The basic idea is that in science, *replication* is the most important element of verifying and validating findings. So if you claim that X causes Y, or that Vitamin C improves disease, or that something causes a problem, what happens is that other scientists that are independent of you will try to investigate that same question and see if they come up with a similar result. If lots of different people come up with the same result and replicate the original finding, then we tend to think that the original finding was probably true and that this is a real relationship or real finding. + +The ultimate standard in strengthening scientific evidence is replication. The goal is to have independent people to do independent things with different data, different methods, and different laboratories and see if you get the same result. There's a sense that if a relationship in nature is truly there, then it should be robust to having different people discover it in different ways. Replication is particularly important in areas where findings can have big policy impacts or can influence regulatory types of decisions. + + + +## What's Wrong with Replication? + +So what's wrong with replication? There's really nothing wrong with it. This is what science has been doing for a long time, through hundreds of years. And there's nothing wrong with it today. But the problem is that it's becoming more and more challenging to do replication or to replicate other studies. Part of the reason is because studies are getting bigger and bigger. + +In order to do big studies you need a lot of money and so, well, there's a lot of money involved! If you want to do ten versions of the same study, you need ten times as much money and there's not as much money around as there used to be. Sometimes it's difficult to replicate a study because if the original study took 20 years to do, it's difficult to wait around another 20 years for replication. Some studies are just plain unique, such as studying the impact of a massive earthquake in a very specific location and time. If you're looking at a unique situation in time or a unique population, you can't readily replicate that situation. + +There are a lot of good reasons why you can't replicate a study. If you can't replicate a study, is the alternative just to do nothing, just let that study stand by itself? The idea behind a reproducible research is to create a kind of minimum standard or a middle ground where we won't be replicating a study, but maybe we can do something in between. The basic problem is that you have the gold standard, which is replication, and then you have the worst standard which is doing nothing. What can we do that's in between the gold standard and doing nothing? That is where reproducibility comes in. That's how we can kind of bridge the gap between replication and nothing. + +In non-research settings, often full replication isn't even the point. Often the goal is to preserve something to the point where anybody in an organization can repeat what you did (for example, after you leave the organization). In this case, reproducibility is key to maintaining the history of a project and making sure that every step along the way is clear. + + + +## Reproducibility to the Rescue + +Why do we need this kind of middle ground? We haven't clearly defined reproducibility yet, but the basic idea is that you need to make the **data** available for the original study and the **computational methods** available so that other people can look at your data and run the kind of analysis that you've run, and come to the same findings that you found. + +What reproducible research is about is a *validation of the data analysis*. Because you're not collecting independent data using independent methods, it's a little bit more difficult to validate the scientific question itself. But if you can take someone's data and reproduce their findings, then you can, in some sense, validate the data analysis. This involves having the data and the code because more likely than not, the analysis will have been done on the computer using some sort of programming language, like R. So you can take their code and their data and reproduce the findings that they come up with. Then you can at least have confidence that the analysis was done appropriately and that the correct methods were used. + +Recently, there's been a lot of discussion of reproducibility in the media and in the scientific literature. The journal *Science* had a special issue on reproducibility and data replication. Other journals have updated policies on publication to encourage reproducibility. In 2012, a feature on the TV show 60 minutes looked at a major incident at Duke University where many results involving a promising cancer test were found to be not reproducible (more on that later). This led to a number of studies and clinical trials having to be stopped, followed by an investigation which is still ongoing. + +## Reproducibility Guidelines + +Finally, the Institute of Medicine (now the National Academy of Medicine), in response to a lot of events involving reproducibility of scientific studies, issued a report saying that best practices should be done to promote and encourage reproducibility, particularly in what's called 'omics based research, such as genomics, proteomics, other similar areas involving high-throughput biological measurements. This was a very important report. Of the many recommendations that the IOM made, the key ones were that + +* Data and metadata need to be made available; + +* Computer code should be fully specified, so that people can examine it to see what was done; + +* All the steps of the computational analysis, including any preprocessing of data, should be fully described so that people can study it and reproduce it. + +We will expand on these ideas in the chapters that follow. + + +## Summary + +* Replication, whereby scientific questions are examined and verified independently by different scientists, is the gold standard for scientific validity. + +* Replication can be difficult and often there are no resources to independently replicate a study. + +* Reproducibility, whereby data and code are re-analyzed by independent scientists to obtain the same results of the original investigator, is a reasonable minimum standard when replication is not possible. + -## Motivation - - -## Target Audience - -The course is intended for ... - -## Curriculum - -The course covers... - - - -```r -devtools::session_info() -``` - -``` -## ─ Session info ─────────────────────────────────────────────────────────────── -## setting value -## version R version 4.0.2 (2020-06-22) -## os Ubuntu 20.04.5 LTS -## system x86_64, linux-gnu -## ui X11 -## language (EN) -## collate en_US.UTF-8 -## ctype en_US.UTF-8 -## tz Etc/UTC -## date 2023-04-11 -## -## ─ Packages ─────────────────────────────────────────────────────────────────── -## package * version date lib source -## assertthat 0.2.1 2019-03-21 [1] RSPM (R 4.0.5) -## bookdown * 0.24 2023-03-28 [1] Github (rstudio/bookdown@88bc4ea) -## cachem 1.0.7 2023-02-24 [1] CRAN (R 4.0.2) -## callr 3.5.0 2020-10-08 [1] RSPM (R 4.0.2) -## cli 3.6.1 2023-03-23 [1] CRAN (R 4.0.2) -## crayon 1.3.4 2017-09-16 [1] RSPM (R 4.0.0) -## desc 1.2.0 2018-05-01 [1] RSPM (R 4.0.3) -## devtools 2.3.2 2020-09-18 [1] RSPM (R 4.0.3) -## digest 0.6.25 2020-02-23 [1] RSPM (R 4.0.0) -## ellipsis 0.3.1 2020-05-15 [1] RSPM (R 4.0.3) -## evaluate 0.20 2023-01-17 [1] CRAN (R 4.0.2) -## fansi 0.4.1 2020-01-08 [1] RSPM (R 4.0.0) -## fastmap 1.1.1 2023-02-24 [1] CRAN (R 4.0.2) -## fs 1.5.0 2020-07-31 [1] RSPM (R 4.0.3) -## glue 1.4.2 2020-08-27 [1] RSPM (R 4.0.5) -## hms 0.5.3 2020-01-08 [1] RSPM (R 4.0.0) -## htmltools 0.5.5 2023-03-23 [1] CRAN (R 4.0.2) -## knitr 1.33 2023-03-28 [1] Github (yihui/knitr@a1052d1) -## lifecycle 1.0.3 2022-10-07 [1] CRAN (R 4.0.2) -## magrittr 2.0.3 2022-03-30 [1] CRAN (R 4.0.2) -## memoise 2.0.1 2021-11-26 [1] CRAN (R 4.0.2) -## ottrpal 1.0.1 2023-03-28 [1] Github (jhudsl/ottrpal@151e412) -## pillar 1.9.0 2023-03-22 [1] CRAN (R 4.0.2) -## pkgbuild 1.1.0 2020-07-13 [1] RSPM (R 4.0.2) -## pkgconfig 2.0.3 2019-09-22 [1] RSPM (R 4.0.3) -## pkgload 1.1.0 2020-05-29 [1] RSPM (R 4.0.3) -## prettyunits 1.1.1 2020-01-24 [1] RSPM (R 4.0.3) -## processx 3.4.4 2020-09-03 [1] RSPM (R 4.0.2) -## ps 1.4.0 2020-10-07 [1] RSPM (R 4.0.2) -## R6 2.4.1 2019-11-12 [1] RSPM (R 4.0.0) -## readr 1.4.0 2020-10-05 [1] RSPM (R 4.0.2) -## remotes 2.2.0 2020-07-21 [1] RSPM (R 4.0.3) -## rlang 1.1.0 2023-03-14 [1] CRAN (R 4.0.2) -## rmarkdown 2.10 2023-03-28 [1] Github (rstudio/rmarkdown@02d3c25) -## rprojroot 2.0.3 2022-04-02 [1] CRAN (R 4.0.2) -## sessioninfo 1.1.1 2018-11-05 [1] RSPM (R 4.0.3) -## stringi 1.5.3 2020-09-09 [1] RSPM (R 4.0.3) -## stringr 1.4.0 2019-02-10 [1] RSPM (R 4.0.3) -## testthat 3.0.1 2023-03-28 [1] Github (R-lib/testthat@e99155a) -## tibble 3.2.1 2023-03-20 [1] CRAN (R 4.0.2) -## usethis 1.6.3 2020-09-17 [1] RSPM (R 4.0.2) -## utf8 1.1.4 2018-05-24 [1] RSPM (R 4.0.3) -## vctrs 0.6.1 2023-03-22 [1] CRAN (R 4.0.2) -## withr 2.3.0 2020-09-22 [1] RSPM (R 4.0.2) -## xfun 0.26 2023-03-28 [1] Github (yihui/xfun@74c2a66) -## yaml 2.2.1 2020-02-01 [1] RSPM (R 4.0.3) -## -## [1] /usr/local/lib/R/site-library -## [2] /usr/local/lib/R/library -``` diff --git a/docs/02-intro.md b/docs/02-intro.md new file mode 100644 index 0000000..4142193 --- /dev/null +++ b/docs/02-intro.md @@ -0,0 +1,67 @@ + + + + +# From "X" to "Computational X" + +What is driving this need for a "reproducibility middle ground" between replication and doing nothing? For starters, there are a lot of new technologies on the scene and in many different fields of study including, biology, chemistry and environmental science. These technologies allow us to collect data at a much higher throughput so we end up with these very complex and very high dimensional data sets. These datasets can be collected almost instantaneously compared to even just ten years ago---the technology has allowed us to create huge data sets at essentially the touch of a button. Furthermore, we the computing power to take existing (already huge) databases and merge them into even bigger and bigger databases. Finally, the massive increase in computing power has allowed us to implement more sophisticated and complex analysis routines. + +The analyses themselves, the models that we fit and the algorithms that we run, are much much more complicated than they used to be. Having a basic understanding of these algorithms is difficult, even for a sophisticated person, and it's almost impossible to describe these algorithms with words alone. Understanding what someone did in a data analysis now requires looking at *code* and scrutinizing the computer programs that people used. + +The bottom line with all these different trends is that for every field "X", there is now "Computational X". There's computational biology, computational astronomy---whatever it is you want, there is a computational version of it. + + + +## Example: Air Pollution and Health + +One example of an area were reproducibility is important comes from the area of air pollution and health. Air pollution and health is a big field and it involves a confluence of features that emphasize the need for reproducibility. + +The first feature is that we are estimating very small, but very important, public health effects in the presence of a numerous much stronger signals. You can think about air pollution as something that's perhaps harmful, but even if it were harmful there are likely many other things that are going to be more harmful that you have to worry about. Pollution is going to be at the very top of the list of things that are going to harm you. In other words, there's an inherently weak signal there. + +Second, the results of a lot of air pollution research inform substantial policy decisions. Many federal air pollution regulations in the United States are based on scientific research in this area and these regulations can affect a lot of stakeholders in government and industry. + +Finally, we use a lot of complex statistical methods to do these studies and these statistical methods are subsequently subjected to intense scrutiny. The combination of an inherently weak signal, substantial policy impacts, and complex statistical methods almost require that the research that we do be reproducible. + + + +## The Data Science Pipeline + + +The basic issue is when you read a description of a data analysis, such as in an article or a technical report, for the most part, what you get is the report and nothing else. Of course, everyone knows that behind the scenes there's *a lot* that went into this report and that's what I call the **data science pipeline**. + +![](resources/images/02-intro_files/figure-docx//1-pVlxwkzTzFH4PHWzm9cuXbKpzlbSd1TvXEy60Sste0_p.png){width=100%} + +In this pipeline, there are two "actors": the *author* of the report/article and the *reader*. On the left side, the author is going from left to right along this pipeline. The reader is going from right to left. If you're the reader you read the article and you want to know more about what happened: Where is the data? What was used here? The basic idea behind reproducibility is to focus on the analytic data and the computational results. With reproducibility the goal is to allow the author of a report and the reader of that report to "meet in the middle". + +## Elements of Reproducibility + +What do we need for reproducibility? There's a variety of ways to talk about this, but one basic definition that we've come up with is that there are four things that are required to make results reproducible: + +1. **Analytic data**. The data that were used for the analysis that was presented should be available for others to access. This is different from the *raw data* because very often in a data analysis the raw data are not all used for the analysis, but rather some subset is used. It may be interesting to see the raw data but impractical to actually have it. Analytic data is key to examining the data analysis. + +2. **Analytic code**. The analytic code is the code that was applied to the analytic data to produce the key results. This may be preprocessing code, regression modeling code, or really any other code used to produce the results from the analytic data. + +3. **Documentation**. Documentation of that code and the data is very important. + +4. **Distribution**. Finally, there needs to be some standard means of distribution, so all this data in the code is easily accessible. + + +## Authors and Readers + +It is important to realize that there are multiple players when you talk about reproducibility--there are different types of parties that have different types of interests. There are authors who produce research and they want to make their research reproducible. There are also readers of research and they want to reproduce that work. Everyone needs tools to make their lives easier. + +One current challenge is that authors of research have to undergo considerable effort to make their results available to a wide audience. Publishing data and code today is not necessarily a trivial task. Although there are a number of resources available now, that were not available even five years ago, it's still a bit of a challenge to get things out on the web (or at least distributed widely). Resources like [GitHub](https://github.com) and various data repositories have made a big difference, but there is still a ways to go with respect to building up the public reproducibility infrastructure. + +Furthermore, even when data and code are available, readers often have to download the data, download the code, and then they have to piece everything together, usually by hand. It's not always an easy task to put the data and code together. Also, readers may not have the same computational resources that the original authors did. If the original authors used an enormous computing cluster, for example, to do their analysis, the readers may not have that same enormous computing cluster at their disposal. It may be difficult for readers to reproduce the same results. + +Generally the toolbox for doing reproducible research is small, although it's definitely growing. In practice, authors often just throw things up on the web. There are journals and supplementary materials, but they are famously disorganized. There are only a few central databases that authors can take advantage of to post their data and make it available. So if you're working in a field that has a central database that everyone uses, that's great. If you're not, then you have to assemble your own resources. We will discuss the tools for doing reproducible research in the third course in this series. + + +## Summary + +* The process of conducting and disseminating research can be depicted as a "data science pipeline" + +* Readers and consumers of data science research are typically not privy to the details of the data science pipeline + +* One view of reproducibility is that it gives research consumers partial access to the raw pipeline elements. + diff --git a/docs/03-learning-from-failure.md b/docs/03-learning-from-failure.md new file mode 100644 index 0000000..49138d2 --- /dev/null +++ b/docs/03-learning-from-failure.md @@ -0,0 +1,13 @@ +# Learning from Failure + +Learning about how data analyses succeeds or fails (but more importantly, fails) is extremely challenging without actually going through the process yourself. I don’t think I ever learned about it except through first hand experience, which took place over the course of years. There are a few reasons for this that I have observed over time: + +* Success in scientific data analysis is usually concerned with whether the claims made based on the results are *true* or not. If the results feel true, and the analysis appears rigorous, then that’s usually the end of the discussion. Focus is put on the result and what should come next. The underlying idea here is not necessarily misguided: Progress in science depends on independent replication, and any given analysis cannot be assigned too much weight. + +* When analyses fail, the results are usually vague and confusing. Furthermore, the public rarely finds out about them because they are not published. This is mostly due to human nature: it’s difficult to motivate oneself to write about an experience that was inconclusive and perhaps incoherent. It can also be embarrassing if honest mistakes were made. Publication of negative studies is a separate matter, because a truly negative study is, in fact, conclusive. But often, we don’t even have that much clarity. + +* In the rare cases where we do find out about data analysis failures, the focus is often on who or what is to blame. In cases where criminal activity has taken place, this is an important aspect. However, identifying who or what is to blame usually doesn’t provide us with generalizable knowledge that we can apply to our own data analyses. The underlying assumption of assigning blame is that this failure was a unique situation that could never have happened if the individual to blame had not been involved. Occasionally, there are cases where there is a clear bug in some software that leads to erroneous results. Fixing the bug in the code will "fix" the results, but even in that situation it’s not always clear that the bug is the ultimate cause of failure (although in this case it is the proximate cause). + +The case study presented in the next chapter is useful for thinking about what kinds of generalizable knowledge we can obtain from data analysis failures. This case is special because it had serious implications and large parts of it played out in public. While we likely will never know *all* of the details, we know enough to have a meaningful discussion about the lessons learned. + + diff --git a/docs/04-case-study-genomic.md b/docs/04-case-study-genomic.md new file mode 100644 index 0000000..ca08f6c --- /dev/null +++ b/docs/04-case-study-genomic.md @@ -0,0 +1,55 @@ +# Case Study: Developing a Genomic Predictor + + +The case study presented here took place largely at Duke University and has at various times been referred to as the "Duke Saga". At a high level, it involved a research group that claimed to have developed a genomic signature that predicted whether a person would respond to cancer chemotherapy. When others tried to reproduce the predictive results of the genomic signature, they were unable to do so. + +A brief time line is given below: + +* In 2006, *Nature Medicine* published a paper by Anil Potti and colleagues titled "Genomic signature to guide the use of chemotherapeutics". The paper claimed to have developed a classifier based on applying microarray technology to cell lines maintained by the National Cancer Institute (NCI). They claimed the classifier could determine which patients would respond to chemotherapy treatment. + +* Keith Baggerly and Kevin Coombes, two biostatisticians at the MD Anderson Cancer Center were inundated by requests from (justifiably) excited colleagues who wanted to use this technology. Baggerly and Coombes attempted to reproduce the results using the published description but were unable to do so. They *were* able to reproduce certain results in the paper after deliberately introducing a series of errors into the data analysis. + +* Since this initial incident, a number of other papers from the same lab were scrutinized and numerous errors in analyses were found, many that one might consider basic data handling and wrangling mistakes. In addition, Baggerly and Coombes found circumstantial evidence of deliberate fraud, such as claiming that certain genes were critical to a classifier even though those genes are not included in the microarray claimed to have been used. + +* Clinical trials were started at Duke where patients were randomized into different arms of the trial based on the flawed techniques developed by Potti. After numerous scientists wrote a letter to NCI director Harold Varmus, Duke suspended the trials to investigate the situation. An internal Duke panel eventually cleared Potti and colleagues of any wrongdoing and restarted the trials. Time passes and eventually it is discovered by the *The Cancer Letter* that Potti lied on an application for federal funding about once being a Rhodes Scholar. Eventually, the trials were stopped, but only after much public scrutiny and a series of lawsuits. + +There is quite a bit more detail to this story, which played out over many years. If you want to hear more about this you can hear about it from [Keith Baggerly himself](https://youtu.be/7gYIs7uYbMo) in this nice lecture. + +This saga has been a difficult one to understand from the perspective of drawing generalizable lessons. While it’s fascinating because of the sheer number of problems that occurred, it's not necessarily clear what intervention could be taken to prevent a similar episode from happening in the future. Problems with reproducibility clearly played a role here given that Baggerly and Coombes were initially unable to reproduce any of the original analyses. + + +## Initial Lessons + + +While the details of the Duke Saga were at time astonishing, it is difficult to draw any conclusion about what actually went wrong and what approach should be taken to prevent something like this from happening again. Most people on the outside were just speculating about what could have happened and the people who really would know the details weren’t talking very much. Here’s what most people seemed to take away from the publicly available information about the saga: + +* **Reproducibility**. There was definitely a reproducible research angle to this saga, in that the analyses that were conducted lacked transparency. There was only sketchy code that was published along with paper and data were not immediately available. However, in a strange sense, much of what came to light did so because the work was ultimately partially reproducible. That is in fact how Baggerly and Coombes discovered all the problems. They were able to reproduce the findings *after* deliberately introducing mistakes in the data. If one went back in time and magically forced everyone in the lab to use R Markdown or Juypter Notebooks, it’s not clear how that would have prevented anything. For starters, everyone within the team had access to the analyses and the data. It’s possible that people outside the team might have discovered problems sooner if the work had been completely reproducible, but Baggerly and Coombes figured things out relatively quickly. Also, that is besides the point: We should not depend on people outside the research team as a primary defense against data analytic failure. It's not clear that reproducibility is one of the lessons learned from this saga because it's not clear that it would have made a difference. + +* **Expertise**. The basic narrative explaining this saga was that the data analyses were poorly done. Statisticians in particular have focused on the use of proprietary software, non-reproducible workflows (like pointing and clicking in Excel), and incorrect application of otherwise sound statistical methodology (e.g. cross validation). Perhaps if better-trained people had been doing the analyses, none of this would have happened. Perhaps genomic analyses are too complicated for the traditionally trained laboratory scientist. The idea is that this kind of work is "hard to do" and that you need better people (or improve existing people). That is the gist of the summary in [this segment from the television show 60 Minutes](https://youtu.be/66dPIFMJ_-A) on the entire saga. + +* **Individual behavior**. Anil Potti, the principal investigator of the study, was eventually fired from Duke over this scandal and most would agree with that decision. If Duke had fired him 10 years ago, then perhaps yes, this research would not have happened at Duke, but it might have happened somewhere else, or it might have happened at Duke but with a different principal investigator. So while Dr. Potti was ultimately responsible for the analyses, his firing does not provide a useful "lesson learned". + +## New Information Appears + +In January 2015, *The Cancer Letter* [published a blockbuster memo](https://cancerletter.com/articles/20150109_1/) written by Bradford Perez, who in 2008 was a medical student trainee in the Potti lab. He saw what was going on in the lab and recognized its shoddiness. Problems that Baggerly and Coombes had to essentially reverse engineer, Perez saw first hand and immediately recognized them as serious. In fact, in 2008 he wrote a memo to the leadership of his institute describing some of those problems: + +> “Fifty-nine cell line samples with mRNA expression data…were split in half to designate sensitive and resistant phenotypes. Then in developing the model, *only those samples which fit the model best in cross validation were included*. Over half of the original samples were removed…. This was an incredibly biased approach which does little more than give the appearance of a successful cross validation.” [emphasis added] + +He further wrote, + +> At this point, I believe that the situation is serious enough that all further analysis should be stopped to evaluate what is known about each predictor and it should be reconsidered which are appropriate to continue using and under what circumstances…. I would argue that at this point nothing…should be taken for granted. All claims of predictor validations should be independently and blindly performed.” + +The memo was ignored by Institute leadership. Nothing was stopped and nothing was changed at the time. Perez eventually took his name off a series of papers and left the lab. + + +## Lessons Learned + +The Perez memo is critical because it fundamentally changes the narrative about what went wrong in this entire saga. Yes, genomic analyses are "hard to do" but clearly there was expertise in the lab to recognize that difficulty and to recognize when statistical methods were being incorrectly applied. The problem was not a lack of training, nor was it simply the result of a few honest data management mistakes here and there. The problem was a breakdown in communication and a total lack of trust between investigators and members of the data analytic team. Perez clearly felt uncomfortable raising these issues in the lab and wrote the memo knowing that he had "much to lose". He thought the problem in the lab was that statistical methods were being misapplied, but the real problem in the lab was that *he did not feel comfortable discussing it*. A breakdown in the relationship between an analyst and an investigator is a serious *data analytic problem*. + +It’s possible to imagine an alternate scenario where a data analyst like Perez sees a problem with the way models are being developed or applied, mentions this to the principal investigator and has a detailed discussion, perhaps seeks outside expertise (e.g. from a statistician), and then modifies the procedure to fix the problem. This is a process that happens pretty much every day in many labs around the world. No data analysis is perfect from start to finish. Changes and course corrections are constantly made along the way. problems that can be traced to data collection can be raised with the principal investigator. When results are given to other investigators, sometimes the results don’t seem right to them and they will seek clarification. Mistakes can be fixed and results can be updated. + +When the relationships between an analyst and various members of the investigator team are strong and there is substantial trust between them, honest mistakes are just minor bumps in the road that can be uncovered, discussed, and fixed. When there is a breakdown in those relationships, the exact same mistakes are covered up, denied, and buried. A breakdown in the relationships between analysts and other investigators on the team generally cannot be fixed with a better statistical method, or a reproducible workflow, or open source software. Recognizing that this is the problem is difficult because often there is no easy solution. + +## Summary + +The data analytic lesson learned from the Duke Saga is that data analysts need to be allowed to say "stop". But also, the ability to do so depends critically on the relationships between the analyst and members of the investigator team. If an analyst feels uncomfortable raising analytic issues with other members, then arguably all analyses done by the team are at risk. No amount of statistical expertise or tooling can fix this fundamental human problem. diff --git a/docs/404.html b/docs/404.html index b1f4ec2..b11606d 100644 --- a/docs/404.html +++ b/docs/404.html @@ -53,92 +53,7 @@ - - - + @@ -159,50 +74,33 @@
  • About this Course
  • 1 Introduction
    • -
  • 2 A new chapter +
  • 2 From “X” to “Computational X”
  • About the Authors
    • -
  • 3 References
    • +
  • References

  • This content was published with bookdown by:

    The Fred Hutch Data Science Lab

    diff --git a/docs/Introduction-to-Reproducible-Research.docx b/docs/Introduction-to-Reproducible-Research.docx new file mode 100644 index 0000000000000000000000000000000000000000..e8252b65b9d9acb959e223d55377b37d058dca27 GIT binary patch literal 180657 zcmZ6yL$D|etZun&+qP}nwr$(CZJlk~wr$(C@&4Pdx~gwalKGcal|ix;q=7+D0DvGM z01R59)ebfY1j7LV0Ia|P0FVIy0JMee?OaUlT=Z2u9Za2d={#(0T2iO1h8Pe;{YN~s zZAfWTlnzZ6z+INWvit&wVsRC(0$FB{?zmgAm4zCX$Tl#lGrjK?GzOD5)Y2EQd=yqq z9yoAE5HSo(m_*V1Y4Vc0JlJw++~9;nYRF6;pcZzR(T1T|;454W%9FI~I2WtvGmZ|E z(fez~i?qqA1sO67ZV=_k;}f1LhtDmgr)+5^j?VwxR24O5aH~@dBu-XWH0p=qP0knT zV0#9n`xKQF+jtF1bvU%ddlX4}kP>D&$Jgm3<8NuuZ-AxU1xP4ETJqn}XR^(u0$7Tc zNG7d@&1wS@CgjsC;hzqXlrBo&FMIpN&3a`J>DXW#oJ$hq@>}E|C39kBA57;Uq4{VI zez5Oqz9}X6-PS$cbb;v=u`^=SRPjZ+n@klS~(aI*FA zUbRynAd&Yx;LhIR%*xHHvZ$M3h-m1#>l>hWpBd8?h#Eg2?G-iQmCh5On*WOK#weRd z0mGGCEqT52w4f%S>lSMvDwK zS_(KaEp#u~`0SNqu(aq8V0%fcLKr$!Rc)aKhKOSAVI#0tm8=Yo<$O9bd8Pc0F_F_q zuhHn%p3(1Qz;KzCR;2o4NWSru>&!$A4M4dIX%0wUCP(g)7C+!{{JaDu_)A|97OV%~cFk)gfc1jJ}534k{O>m+LPAQAE z)_iMac9x1=^L{9(d^+b^?q@b;s)(B0$xqjK#uWAIV4suoHY%Ccd4}d43>_0I>uf)a zuBoXs8qM75Za1cm5x85Ky-07**f)r=Q}id*>jNNyg*t)6fA-ziM*8QWPp^zjo0mU) zfez2N&Q-GE?*G@N`v>wDpLplV8b5X#4gR#KQp&>!@At=5H$`MQ$|2RQ^c0=%OWQfz zu|Pvd{I?Z;k^_d5Nl5`nJDB$J7!B^afuUPDQ`Pxx zAU6lJ|5bEeC;GhSZx{8s_xkzuDE+*q`Mf5-ppPA}FU|oTFgbm+Xtjl6iY(FId! zoCJ|P4{4Yry<45$T5Tm%<)JnDv|)+>tUTFS7%ICoPIbZz5*w-1dD=l))Va?u?r1vp z?7_$O)Ct6wnzz2bygEfOi|muO$=D>wqZCVP+d(Rak}AdSRtV36C78GowSW{%jRGRi z8-3N1NZueD87gwb5ibU!D@iR0pn-NaMbVHPNl7S^BC5C~ENe;?W@d6uLp18|wN3^z zg~YsQaxm2Gll2sm_fS$wV5~k-J0(qX>4Nm zsKGBK!%BCd>V+UJW1(xwXigb3YIZ?LH|4QWHnQ|8C#yFfu4r23858>TBY~n(e-{)+ z)UAE!a|Xh04}D(=YHC*2l=4el%^4}J&C*;rtbu z*gobVlAX!gariPj`qhz`0;8og*cNGUj$V{+aq&8=f=)^{B)xrM@0xY)%|D}?n2dk^ z-k|)nyc}fglGpqqD^-T-?KH?9O4qPEIVY6Pp80LGl)MbcIp0%6uTOqE)HBe(r0aWS zl;BT_E^Oy#>1Ss%@lF7t55|<**SH1nYArqoWK&utI_#jc^n**gMhCCU-l_H5$lJ_q zc82edk8UZI6X#*A__$-i!0CCK#jN-jS11;JgzaAl~;;l3ajpDyyYUv;s$# z#6yE&B~{%Vl_ommgujzAzUqM&(GijECw)UUp~ZMhX~-H-7ksnAnyb-7 zu-H}q0wd(Bl`xCqgrY#&?FKvICT1lkxj5~p!sL`*mwd1b%r8{|eQ{XpJys=VKbQR& z?-HJ5lqe3(TzQ_LEl6k!#+pNI9O za^(POi9%|IS9-)y6;o?mK%Xx-stQGjSdB#Ko~-SI*j*{kuh% zReg6Mmg|1T5RP6|RIQIVQ8KnEdr>oCb;`|*3$Qbt4bhmU9r%Z&<=ppxWXln2OXl^K ze)jeIt|x+$$@jIW<|u~k1wuL1Wm+{wqZ2dLGJim}h}g)w4x=+yYX|v?QPw%zWVt^v5_WWfvm%6Hyy@%aD(L#O2}zKiUM6 ze2p|!&f;=^w_fK&Qisgv&T6xtay9k^#19P^9fXdjhOiyaDZC6+5Pq0o6CvM)QD82p zS1M{2J;pW_N6y=Yq4FiDFECY9DRQ9uXI&eu=F?SBHbl) z57dt~%T8`E5|=$>kHuOV?Oe1|-9ZWfK^{;;_I+E5 zxg^kX%z~FBa3Yq`I65Kg5@|MkNX-IBxBN4@alFhI$CzIr#u~HNP9|crqZZ>OF>PGq z39$;r1QzaQstc(mVwE0;A{s5lV5kp90cSacA;N<}`FS5yyXs+BO=uMtKOfO6)&N-! z>XZpI8ptEQ9E7;9{E8U~LMoZzlSZq4ak2SQHD$qO4?cZBWD!6c~W zxA$d+bq999HNk{D!Mk%(+hh3wM$W3~bN~mt$CY4|q!YtgjuZ1@h*?UpB_y^%`^wX8Rq^kD`BB=5J}O-HOB`nqhRaw(Big#FfXE^A*Z|{j zOFi9?mH(PAWZ$DIU|)V(@N;r(>(g6CX|_5{T}2EI)BNb0;vx~-XkQ$ zU9sd6oJT7Xh+#?g2IUPBw_&pc z49DLv*@7^Fv@;elQ}DboP!W8ubgXck1IFU3{+Shl$%ymI{du%;Q_4AYQ1q{JvC3gK z$uvMj3G=vFnk>)aGy3OrD(o;_h=VV06&sA!Q{b>sz7mCCz#9g)OnHNCy9GVJg;a9 zR)s8?rf^qDMQGt_wyk0f1jsNC+ZxMsxHz|ySj9pp98%T;t!7cDT`KINs*i2_Fe_Tz zeqE_sAifns`%7+l)couTjZ_3}U@>KeVkZW{V3im}-Oz;H`9BFq5wr_N{4euuS);@) z>leHmGJ!MheSZ9|h6d`xnDx7dsh@SD+zVwQf6fHTn66F`g<`)-sD%b4V!0WKV@r#? zL(MF!7i0T)rJcf-^kl7pO)K#&h4WB+uu~TMTNFTL7UL^XEl5|uF)~&u7mRodF9n`r zTHf7ft&142^9IKFRCJ6kFI@pH?sFKRLUOg)6jYw{@-H$*J;#VTL+epfwDcda!!YElax2Zj_HlA6fS1!fw-QIvDd-^iMN^Bf5dc%j4$+5j zjLga*cZY$!OvVG!anYqlQnbJbnZ;|1| z)-ykeQSS!J9i-@t5rsbuTXVuLD~vk!5hW%fw`<5fX7+`h3Xrxe&%s^k*8Ra*fufo0 zMD(}F7=n!k!he>sXNDDC9P8LbXkJFc4mEApL-7o%^o3)<=lM9V@VlVWN;7Ja`4Jmi zKGTpe@7GmfcFXV`rpXAjf-1mrVF=8fMM#BlNCcTR!=@rzu?ke3bTRAGox}4(EUiPj zcuWrl1dbC30{0S;3RFRAI43j*PPO68y#bb`Y-c>-ScKrT6^EV@k{NTb7j?-%3t>o= z(_{{3_Jd|#u&I1RpCVF1j-_dBg7h9fd3yK9Ha!NL8ggjuW*)yYl%QYOjK7ys1MGf* z?j?xbAwHENv37x_tpKLG1gRkck$9&ezwe*imo>YY5ug8Zb>IfckX zKVjo<0T5J`uOSrWi4Zn5Q7d%kK{{s;Gh(w6`!YswgHYays4e$R%y+)_BkCz_w}Ff? zCz2w4#B-%lBsZgxE;SuW$c_*?SRHJiU0QQ1Q&e_kZa?E47gf{&92}6m5jn@ynG4CW zWE|(VVEQW5=2z26w1Evm8Bq%8#e)eUq;~CT*1f;Ku)K*8Pq8<^&61Nm5A)`ZH%bqB zU+U604vSz7uNjoROX!HnSi($>QX_q~z}@0t~Z>gyaCWBcFN%%K(d#wx_#Q3oC@V$6?w0a@it-G->szEf@QA(ba9O;#m)7NwV4` zP)#-JAPU<0YV%#46(Yjr(-$}l5iAOruvtnQvO~?!wUATY4Cp>ic<4TxP6L$_=T|3b z_1(eNjoojJS=l2G!PE!uUGq$HMpzSE*vrxnnin;w(OlqY3qNozT7wsg8TXrFFCO$# zYb@~0iVV2a=?P+OHugH*u2o4-p$=wFu7W4`Y*HN7q<~Tfqt#t`ttN@Oj{SszYrdl7 zX7rNz7dzyM5yHk|7(J>YBM@qDEAyE0LOqi8sO8a(XU17nh1ChSYdX5o8&U3}#vNy% ziJD$VgDPhi8-~4evxE{~ff{kcC%=@{&$wap{5w$@dskSO`iqx#k{3TO>3^<~Wq(c; ziGOdj13uO#3UqCl;0nyWNpP9*bcHpi8ZD~?|_prg4b(s~a@ zeIRcjp-iSL&=`qp+|ErFPm9Y$Bw*~x*ye%SluCj;ys7mxsRFqsTgtSt86Z^D2rcyK zl{$IrjpP3s1{R3LnA!FOVd2Woo@dGv*TERmsf!GSkx-;8!WaNsIt@p5xkP5c2-{hh zu7Z>1{Vz$7DozEe2o0z@D0uu#MKXuI-vRu&OKSrJswGL|xWMQB`O7Y-R7?wQO;%!b zO_F{Bv+2RTMwX4#^?jqiw#)z?bF^#Rh|6M3Gg_CHf3*UI80AUa4T`DTC zF}(HINa4`K=;2uJRT2qAaLy=whG;lTi3i8yk5US`r16{(=OJ_Xlt`IT4nSjGR-O;8 z60L5gLxj!C9SS8vaQ8QHB3PmpywKKj(8S4@FX^gIA0(SsP&n?=@XE19BBL!cfbA&K zG|_to#a6(4V|3~*pDU1>1ELGu{9rnAktUYrncl1&<_6Cs3$K&J)ua$hyLMdP7@P@W zOutg86_vLqL=r6Qgn<-C6&Lt1l)w%ydO1ZQIdAN2bHW}K+UJZ>zGwmy2{;NUEA$j_ zBku8_a8+@T1sVMO1Ku7K{TuoYX@`T8y;@}pz?5Bzts+3U@vHtqu(6NybCx)12f5so zpM^ZlRlUrtS{X9zU?tw;6wek~Rr zlL0ktGiqQb|M8ve*a`D@kJY7pyW8xSfTMUM(nf*@fn}hR{`aUjIN&x)ZEwBIAhD&H zRukK)q0lKIJoec@`{jIi0vkrA%6c&;le5}XQ;!8(mn(SRLi6PLJeani-6ltNUgweP zPI~B{kBwe?M)jHiLD2dPJK#|>!dLz=;)85&2%t6B-Gs;4hO2i`Y7df#ur)g@k*S64 ziO`g-`=n|C_F_&&$-I$6f~-C>dPPmqZw@2A+Mdm;XuCmx6DJ$frI!6%|ERf3| zF%qN61J2wkQx+?9>BL#f_53nDu(5CezdAHxNdmdk2amv8Ef>G|9A+_6fQ}$ffE;j~ zdjR*JE$o;=K7zfW?sTp$6{)49AHT7eO!URNI$c2{&$47zK)%N)O$u5I!$X5zHIFm6 zhUjp`cyiupg~wF(8f2IFhM2_(MKF@;OhhVv2#E!mM)o#p&j6t;Or$TtH#M*E}PHOq# z*4mlg1xzsr9Co`1(iWv(LdV|85XG2d+oa_ek1YiB%x#^ABHi}*tV}T)WuEw zRwcl~vVT6$D|id}TbyI4Hjx|-y4bXGsl20D#V9>1FFUCi^QR#a=Se2Y$6ADu#@Qm7 zIo{P*R>8I)fh5olI6{>b;x>|o$4LDCbK)q>QG^!%)=Cixl+}Nvu_zHu^hAo=vEz^v zqKJ4EO$ti!&dSM34^WtC(}`9aDoHQjit?CV+G(R(w}s^((_EgZwmIe&J%ja~NydLh ztG=EWKgekQV#J(-Q}nNTm+GF**uejbl{a<j7VJx&1v~`L zc=e@)hsQ`gkKub0c?>+-1VdZYbg>U{1={^H!X@$KN4UY{NxepkH(^V_zU zSzOSu+Na*~1P%p|2CRM$?Z~+NR=g?0WDDI?6TvbnS?YDxciL6rofyZ(J!7Y3CCP`^ z-*O2>dH=24$1Y-DTCwA|)Q*FcWs7hBj0-ETlK);f=X@1xaezr+0AQxAY8 zlYFLV1RXvP7Y-i3+AowA)clnHa=r9(jBa<)Rn56CJ4U1n_Za4HRM(n*#m-^PPSM1< z!X&ObL@&jw0oRW|sVC}@N#@)MEdPvMVuOC_5Gx70PTBFk+uXO3Z+){Cna++4@M7^K zK!Ng6X|(-LG*xN_Byxo=K#{F>#d@Q3aE{<8_e!I)djz0E6m{Myb;tb#*9$Y)kRds>?1 zs2GENx8)-GqxLu_L`+;Y+PyGd>#Hz3ZGFZYHQ%)UYNQx;-1X=RjB$2^j_m~u%5w(- za3bb`)oEnS*L(FSy3%pzX}P@suTibB@^hrP&q+bMa9o!i;MZE0FyU;66LsBAG4%)< zgJi^xA>40+(kEo{L~L+X(O*Jm?xLY?0sI7RmH%ABDVo!E)oMUZjgKV0L0N066>)1a zn0!1|hjqNd_n&Nut6A-Q!TonrwtK~S=g?EpRc+};H-zPisV|Grz6-OiBuW{Hdy`m6 z4u(xX15%6P_>PNis|C~qHnB?w?L9I`c=^L`-n)5s`?Z=b-KCfl(gp*~KrXh=y38#q z0_<1yMsVi?z%gD*gNd+%ACZi^O|3z?{c~2d!#;;WCw1(worQ{zITM{>w&f%^%_K)f z=RrFnWx85XgwpumrV#{ES65O(MPzngNtKgiU5`#GhtC!y{&!a&K1o6GDhXX`C$*Ra zTdX|5NSYUB5Pf8x0sAdJasl5+GfNWE2AkJ`eX|a}xd+!CIPeLO&>$ynto4pLVMt#n zvC9qwy*8XP`DOP-Xf=c8nz8p9{cH9*8pz^?qVt6=`$u^Vrd<^f%i@LdGqDD1Y0nFA&Ej}`7ooCEsu4{~6qk$#WF{8r)$N4Hzl7_=s4 z(HXbqhAyOY-Bs;cu64mk#d{a`VCr21qx=!6Mqk?=_~3RF&ifVpUlA@r1e#MS0cZBS z=Vm6Qt;aZHr&O>kiG^yL4Wl(=zB|qZtG*X4vvu<8k1(#k5{Hu(YUMvm?HP;6&9Od@NKa$YF9X?IPx5{?ZKJB zSm=W}8M4Bf+H0z(Y&EgtkX2Bu@ev1QrdjsBO3y@@vs?#LwySR03QIsq=+Sxv?`~Hg(jX*4%DxV@bq$0Sn~dq}2DH69 zJX$#^-lnQW7FOo+8X1bu2y~R{qXWo_cEv(4z8?@T>bMyY=IvoVo&WZKe+OpJw)@Oj zf2KwS-8yv!7?d<=+omKaBWrM`m)O!CQ&Wajj%Q~svO@Z+N`}Q_WjSF|T_oW=6Fo9s zpiq*L0IgxW;BSe@2w=9&woQn-@^nua+tCxK$_Z{S3y);AhHKiFuKY-tO6qFITxNQ> z+gyN0ZZ4?W5L=OTc#~K~12GORa$XW!!OSPQ<9v47wH`xH7MdqlE@RUP9ZMADmTQ0t z8Nr@PemTf(W(dv}PgkvsxSi70LtvA!`%EG+9Zo@ST$Mo^&X1Zatg4%p&1S=O*O}&5 zw3K(TwrhLeH<&GS5HCm9qvmDVa2j~{9ttGHM<|6U zD#*;$oEH_E*A?y6300?gAkMSWMuM!_dS`tzgY^1ytnwuosWFETLT3I6r!3e7U=v@`% z6V()3H@dT?Jjyxf=%|tnaXJk!KF>X_*^xVS>}vIo6$!t1nj{<*K2am@&B*Rpg_yop zZ^y!V4S9|rz2VoD1i!ttifD6mm$o`kLmtG}rNDm+Mp+5#C6Bk{c=uU+K@7qT62g%f z`zW9|Y%FY(tvh6%-i$wY*`1GzB9Y(iE2spk=w+)z?AvA|cfGV&!FIb%aoU6^b?3@0 z=e>JnM+VmImKpo|#ot=4%sb83I~bEP>fiBZHQufq7VhDzi4qdm0}j>Tt*W$1t&yJ` zqFA-qDBrJmw$Qxh+nqO0a~^Nfdr>cr)?H0J&g5>0eJH89shh#`cVy5i)KzZ{cUwWg zw;u>h=yp~`cn>y|ZE8y!$Q{w+H-KI@ixCnP2Y}@Q>%7U>d#?dF+S(|9qqt05BbZ|9 z7uCX|C#BYZu1(LiT)7-Z86&0hpRB6cF-XE^IHuiP;B2R~5L5DvmRz1$#`OOpD_5cs z$S(OPhxWq^2E|py+-_B`ARUb<2{9h-$hh!i?=)lIOr%l_e2LZ8-+;uq?=W437#8HE}R!sTmXxD3VH*!HdDqTseBm_tLt2bdcsrlFQ2UQ_Jrr8^}z~gm$m4-;xco z8t@YGRteNf%u%LNxpdZyN{d0QFwsAp;JX*5@OC$DBGwbCY&=&I->UW3eDv_`gt z#i~%Ma~bW{mySDDmc@aE7P5DA_+B3UqC z>J*~_9c|z$OAJmA5~!Vq@j|P2q2rt5S&*n9?p)RoHSt#@dsNyMCt#1fD5)<_3O7kk z^V%t%tHc=ge(0T2w|TKfgTgcbM|qL9Av$i(_h6~K)AKi-DgSad` zvN+Q)SHCJ7^!-(+8fxG_`q+*l$=ysVolO}SA(EFI0ih5o+8f7A;Z|SZ#IM6#Vn%EP z_d@j%pzn5cOgy=70R!R#i4!SWNLXB};vWsXu- zRwQ$RrUIjyc~Q971U~`pg*Rg^D7NF4-Emc&`_^p@stfTVv(2=$(kH!PI%5jaVzzLa z6w?7#4lVO9Y41xR3>SdAhL8^<>Xg|?-kz{1PiF=29D?Z)6%Yum-)c;slaOx8GP)69chH!U;0;7a48o zQ5-Zu^XMkH$+R}xFy$ojhPBEf-F2~gEj>CG5F8M->Aq`NygPnRuX__G2CqjF&orJL z`8-}vE}w4`pZE8TeFqjlo7YjY4XCYtS74&Jfa5zyz;+0^@AaTqS zFH+onzVV>@Dz&y#Z!qxzFGFWKNd#v^^0U(vITzxJT~4b`#NXj)4}5ug z@Q}Ag?J+0V@;tC) z2Nt_qBeM32P2BcJy#m6_Oo=bPRy(4R;FnEA+vCG9T2A@8N-Bb$`@Bs@fAaPtsMftj z+viS(dPXvz5c?ldyQOnm#Wgk8`;>Jk%j48yH#>2`68qIUO~i&s(GdQ`_aJJMtj#!k zdREyKf{mf4abREZX>X&3mD%#KOVqZ2mZQdi&OVr_r$8c%q~P|YfIAfORNJfi`M<>Z z*jD=fp}ROe>8TG>=#&gJT!ccML=YpXS|_^GO-|T}NW80NN&+d0Zq&7VsqK0@{}$*k zb70@sc*ZC|E9xx%Y~<$vdZ;A4f_D0Ov)20;0-7LMRcDVU)T~b^(bZTUpk&%miOueb z64MEb2~FGxNR-GALLp!a_65Kn#OgyGDUxTX0DmSS;x zqL;W{X=XxuxT7kZG^PN$mc(&prEPdiV4gM+;ccv~nL=liozmwS9h0{ALMdQzkk0t| zi7^NDNUd)7Mcw0oI*GEAvh4m zPn?i92EIY1^*7=VF$m=Y*-mfp8UB^!|&*s(NP!OJpNmPy7UJc3-?*6=@R5sxtP@Q`Ahs%)l zF|E^n)8%uS2xR|5*5K}UjLY`vZh*-mw;hdS7Cnb-TNQagUf6e%1r0O-39d8h%LV%K z-Cl|gW0*~hH6^9Ac3j!`BtTKjo+#;*yT8PO0Qrc6x~wQ&YDws1wLbTi9oEK~wotM> z0>b_MI-XIUSy?Jwr{z%alI*Z;3Pwl3q7s&{^l=i)!~?h1&|nIt;sZ{h{9y~EX~I&< z^51(muT_8kXY8Ojr*V7FJNp?52@mqF$qTvAl-TB*kw1=B8crum!XiT zzHHQCPjcG6RY|5r-oeXg_{i9H?5|c z*|jz2%_xNBU~Bbs-_ViTL?t-PPZADK85Z{RkfJ>EeL)1kT9o{b zr%LZ=KH_%xjl)fI8iD2kiftEGG0EkP%<{c7-V?FLL%SOy{s9}wQO_aM%i`8HD02`n z6E{u7@ZMl72gj+}Xvtc=HQFm77Z-{0zAPS^(-7me!g*T~vxSHz4&GN*lL6=dR%vY- z@LW1*z1nEG0PVx=xz`>N!|6(UDBqscWtdeo)EzgVXSZSSQPY-0FK^xcD5#C{r3ri^ z-XU}0@xPc%HK&4`3r}?$Y|oQLI1&A%b!%YxJ&YYSYb9tZm%{O}DF?5#@=9YBbnRM` z7VuVyTJT{MjTe_S^;PnPjTr8yNv_T9P&PKAvsG0E8ei^n8EsXF-O+{tg!t|epcz%Dt~b=QV#2tL}8KF491f~A8q-!lzpay=sZIna2y(3sRkAC6Ehf*?1DOON!zsmi4L_=ERx!7WL<+maPW2_f4EzbVrNjs^|EVrBD5XDh)AwUI#}foOp_OV z&j4!O$sqRO@HT!anpAQ?dYX^B#A`GL5dnIQ0j3LbhQuknn_jbm2WB>{S?Ha!LQlFQ z&dW|is=k8_X$r@J&tqq@y?UU(YBZcu3ISTA74iI_hre90iMW|uer`&vGH*c`0kM4a z2isWmFJdkcVGN~s$u0sBy$3_@-*GfeS49YN9xzV)l@j*~_8~KgPr1e4u>LQ^P3qA+%V(qIU3Fwg4F*W$HAw0IpDlI}azS%uD|lO&R*BOUjo;ZSdmDvKt7iE190kY2y^ zF7*HY!M;<05N96q-$eoU28ky<>4x$>-|oxxDYWF&;P#xe3N_R?X6EG9XN-Flo;s_p z)25c<+2so@cb2oX3VA7ySq4*IGoc+IJLzEphh%k;K(8@1;DVE=2~Be3>{{U>j*22x za7Kk+fhBQmW5gy7Q(T{`ef7m-6T_H@_}?_gA#IE`<~c~~#FKe>tXzQs=EPyJ8E9}R zr+3Yfk+;R5wGHt?F7~97B4W=Wb70FrLdJx3OIW63PqP@x4S!57bDa0n2Y<*{81uUJ$QnI}1(K^%Ocr(Mob zKYBSJta2M7+v~3+(k7nP>0gP!tnJw;G}>XN9_L{GOW-{(qzY~p46z;2!KF>zBn_c4 zYOuN)3GCu@2X*X&HF=dq^;-$)KR%@QVbq8S>D?7dbt3oJpY1GL#JQJ^3JH+QbN{YT2J z(Jb(TQIpy~8xoUXIEk8~RaR)?ADv&ra@CYfq?)Di9h~NCteMO`)jRyV0!nBHKtMz) zR;p*Q+R=6a@h~|bT6LY)^f*h&(S&2_QpUvCoEPRsCA^{S2EByWp5Z?`Qj@jHd+9>B zKxrjDHK6U={1j7TrtOHkP6&_h<49SyXo~EthZ(zGD;)0KZod2WjG03^t*G*Hg3Z}@ z(#?1$KC%zQZnSf}$w>&-0b`^yI29mXDElTf{@8Lro1z;2(J%#9-eM)YCRH6Oq2bq^ ze{-?R!`oxMg>2utvsF2kuR&pi?KHHN?Ihzxiz|>gS2ggFi$>5A;#aXVi1Cm@9f;h- zW$~+oB6s6Gov|BsC=B4vkuTY8TJj-1;#mujZ(psbM+T|1=Do`xuYjw*+qL(I_Y&6F zMi4g`QzI^TVi1JwiP>v4v`&%*#nDi9U~6yvGu(Tqx#RYAn8)l&OK6T%i|D9$uM<>s zA!O{n?q4N{Yzg$?-qVOa?B)@O=>!Le=zsv#j*MFoKx_%5^Cz=Wuq>6JQ7LuV_1}JU z6F*Ac)P&4jaY`G;94pjdfE)^f8sjCDG-UepK-r#}fSn^JQtVvpmlHhw{%TIDcwg+7z>^l-aR<+n_< zf@`FnD|(~j_5;;)rvnG6bKnAuzpnX(bSS8ORVxxmdtH+n5TzSoW=4n}n`LuL8l)s; z*Omcf7<3fqb<1_A4)-%EMh&)X3z>r&_20+H7QwH8b5uL~Ea6rENl<(i=Zv}26&W@w zDYQ@i0mT8?6Wr9yMnE*(Vz9g0rL<_<$^uCIGH99Jv#Z-)a@s8Rkl;Iab5U$DN&`Hh zpsX*VZ%@hz6^AQG90Nc*Y)swI&n$F>z8>mg)$btH?SLoM?SU8X;XxWj+ZyevP zd-<|kXm~An6RfgD(V*_U^mqn!` zS;fReFHRvd#=I!qEHPZoQ`c^>c}=1o909~Qo5=Ac4DS?vbETGyUyV;(;}8qks1QfxF}&w`=QPWR{NXwP^21-OW~}j zFynh^l| zHJqPQ)2#u$&gmi`7>yD_cC?8DZ6tM9=$*jn=h&WoBwQ`bbwHhX#qYc=Zb>ewPya5Z z+mbu}g);|+<*{SzSA9YBK+nt$^4?%!kH>7@q5DXXcN=|PS;H@@$9^vlky`%ACuHxP zPXc<~uLQ}=-;42mdPpK{7;sx1;qzV!+d!T2rz9~{1bBQn-0+PjbN(c=C)g#&3|Ho4$3#;*RowEQa@x=9hQ` zES_*W7ojE>Ke+AOUi%fumyO2h0W5eI{)nli>~rFcskS@&oDnK5T(6)mQ?j!|{Lq-c zoy&G*@}mh;F?I$-i%aW4&g~W74&}o8-#CQ{F&=&qeXiPCXvOIPD8=pWjGdn1+KG>@ zT|0vv&!;~q`)!*2uEX|Mev~F&k7ebu1H^s>F7b#$ z!JUJH2U|#Io$JXbJb@Q2E3a85W!y7&v$MLL2_>-kF{1R3#gxtncUZ!yX>~ zb3(~l&TE;cb6xnlM(!uQx3{^$s&7E*=tpxP)Ytvh9iA<`2E=!;EeF%bo}WkG(@=Q& zPuUg#nodcES66=}|HZQY+R+X^ACBz=A(>CmGkjvj&+mRgALP>=PBnKvf{rGv=!$(d z@gBWdBNJ+AZXP;AQkEui=^ig@#59P&Q+jV)P!KGXi*y4Pw!1wu1^xbwGe4EJ_BC*Ej$D2ExI~dS=L7#Zi z?h>oY23Ef^_Hz43ogD5Q4>eH%MSXB`=KSSUB8R!_gH-FKfUS$U2Q=CJhwTy z)NBe|y4&iyuHOm#!|uLA#x-vzeSecj6p}vF7lH&$06gIPuYt8Yj^w>Wi3Ol~rBJg& z9luX^OTb)wSmb#}-@z4NT(InIv&v~WY6^w;8fd3muj=w~J_Y;U>YR|!HR7?4-TJ|I` z-?c(2tXrz>zyp%n8-Oa0s^M%$ThF3lW(ON`sUehK*Y zf}@6YxX5H1ngdbj?l{aW=XutCXZn*ae@l_>+a&&Uleu?z_(PF5V0Q(KIZzj_|F&Ri z&&^4W#6LS$0*yv6!2KcD*$&!(e7UUd-28nb#C@F=*t+k7i2)A`e!uVYT>$-2%}pS_ z=UNoT0o%HTr&;o0zH2tsj%>6ej-3%x|2FN7&Hdf66?Y4;`BF#s7Nk%&bfP=R`^#5&-67@9i3IAu( z#=rsq;QYV#=>Jt}|Gy6D|D)H=q|4i93m}a2K^0yu(7gLsawwLNRc>`wD!nBnx6lAr z19w@t*B{yM4LHJ7C@w=Ijhzf+@5Q+1AAG)Y&gI}@P`7Bc2cC7oQT!vWNDBQj_vO#% zYGYr6PjD>ewxNQ)nkSy5+eT@4AI?CfJl^)|1Q3;~mYq#ZBQpd=m*)Xwd;XzhmW6+- zDQn_Gx#y4LiYp1A*Gl1}uwgW-oEByXx#OC$ILIki5z=gg?0RXB=e}6TZzNS%q5Tnckmftl*qAT0@bTTLoFl z+NbCq7kn$4FK1zUJ+0=9N?vCTo%bYNL3qMCRK3a)JVlXx`LHM!flEBT_E?~1wI~oJbQ~5{=3>4+PNA@jVpiNVGdI4Ui#1}jKyYDQ>|AKav(it;~F>M z`MkhjrvN3z(F8;0zprBhyaN6nwhv|(8{8cy86mhwinPS^QhEdOVCec~Xu0+)f|68`NLqUJGku zw6C#<`J^=61RuG*FtsXx@#!thEx?QLd}(Z=QOMScKGS0IESRq0MOc5q^4Zd@@4E%V z_ZH~7+-N*o-E{0|X?|kdzIa#P$eypt?hp9?gYN=et@l3#e-{D(0QY~xXJ&8jVrTDS z>ij`mD_a5)`_sSm5R(>wuhFWNSFISKOu@?Rh7vC6N zTeoZ*+qP}nwr$(CZ96-*tsQ&Ewr%e?c{%64bKd*zY;CR9)|kKM=(TG0UNx%M2YidS zXucJ)nKrB7SMWg2DF;AL2OYjivsgG5GYhqsxZrMSDTf+ z72rsRY~`~7?j`=csb;d}C$Ylvz5f+-c6q;vB0W!vrkK!k#|Bf*BEhCiqCbi6QI#F5zjaqjP; zIOy$g;YZ?5O|Q*hys|?#MrHIm9=XVdfPF2RAWEC*3T+Gf9la!H29q_lA zqp8q3eSKX!^hOlwQmZO8@G+kT*hg7K6KJ7(i$Rf$nfaLUw@W%zBARpc7P1YM;Y=yw zv=i6j-60m3^azGTuOiIad!bOpyBd`B@eNy?(}#{Zu5q5oiv_B~z^@KQt769JB=m80 zHO8Kg)Q0JTmiv)T=L6KbnWQ(wQ|@l1&4@y1jntf`{oUUt&+p7v246_7V$;`L&zIp= zc#$?QSl;>GW>amtHbNbC$QJMNzfDeLwOURLTznX;HV~MceVCFv5tyujn0CSw)((<2 zYXJ$JN64FK6+qj6yc)pDfqp}d@hE`493*%A%|sQDTYtvv-YRXb>tZ)E@t=uSxbk-{ zbIU5p5RHWJ^v~j(TWj{khqVC;P<#N8wE>J!bOw^O5sqMd0G7Q3`1u7YdkGYw_y8t* z`LBBrvX}q52Pw+~ju6k@=2eE~WtPi7gpdFAd#UE|ZtppJOIpA(+kzSJ$hL3`B&98~ z69*{uBVRs*0i&s4}?s>l?E_Tpm^r87|*jWEO zM^UB@RolCAa&c#69+F&>VsgdAaK^;2#>DW(#4yLiaL2^3$HaX1GSTnoJ+y2{e>*%G`c@7&twy%q{8vGXRSdIH96`vd*Il008`d0?^sT z)8>B?&Xta~`#J}L-;RF&UiT4_1VPyu9>Ecdd6#VK>|f1QpvBcR^|WL&|4*rVS+vFm z2HAbQ^>DQ;OwXQ7d3Nt4cmFS!++|=Az=3VmuIJVbJHeR&wtF`9tGbETY{mDP=NmkF z;J3rR{Nd*A-6Wlswv@N@TsZ^V0_dQ&)o{Gznx7V!schrf?h}^gEtYlxH1GPEd)Emr z5U#*aC5QGt*o|l1ICTz>vB5Wp76W9GbtQ7d8SsAn=G&qrBb=?Zb`L8A4%Q?7l-Ql* zI=VAyx)`C=^IF(^y87(AT0|2IqGb%jGbjd5Atb=sWmQNyO zZ(xJ!?!kcYP(vk<`e0BFVntUK= zZVY?lhryE`!M5UP;^Y{#o#@&IP2e)&XMQeh7;<9CAI^>43Pj7(a^(qemq~J?9ir`p zV$BmUP?>Fmlj|oy%J!RATBpb9#e4)se597 zVu7{OZd3&#$Jx>`SxopjE{;o&G}~TqR9NobNRjL6vOV4{*+r?r4G~Yu&dtYj<_hlU zF+{R~D@2o)^=$A7chgt&Om!LrtK3zTLMb{dNwf}5_H|w%LBdmVZ}!T-$sMbyNVB~m z5E;N96z#oqJq}OPZR|gT%`|1Idl-@5XU2yZ+&_IGqp0j*Tu`oZDgNxUP_03Txl(wU z$;>q&i=syTl}Az4Q?5k8T`V~5%P0)~*V-@Ez{2td6kIH27mvbnp*cZfigf~|U)ag_ zA0kqrDT8DQ_pHgc0~sOcWGeQNTlDOrDgZfo@~f%c>^SEL%_robvJEqo(cp%qB?TRy68zXQHqjtN8pjcZIz5f5w{(e zvW$#`kxaT@eIT^;WT_k2Tm@L;xNh$6;)r$Fs-$|pnmA*5h9&~j6WW(12$Y)COx$gNQ0UYRM`Qx~n&!n?S-g z*HLYdb&+C&lq3h*?0(KWPA9bnFJ6NxH6t^m#8LJldV3g%7h3hDnpbX;jP>z3AW{~0 z7-+!^oD7;EAFHI`->3*c!kqb3MM*T)ew^FR0g&~3H?-cF%ATFAX{`!b!p`dTy6LL$ zCgWc3u~fro*C_U|gW@~V!w1+9lpF6g)>;o1&%MT^T&00>esALv2^l{tgI}MEeXkCD z*H?!5L;R^0-e^2HNYR|eP4C^Em5_?kW2*}jaA1F0Q+8BBGRPA@2mK50xSgrzhAv#I z5p1rC7n}^5F>4YBvn~+7sTYmdO#0j%%%jCQuX}fqpG_lg+#RHeXl-D*kP4)caIPe` z05Nnk@}p}-C4=Gt?y(f=P0+Yf8ZSK!@euKwdjQ3w=~5T_mrea?l!NdfH&Q%5G{I?6 zSy0zD`vcAsN`gv>4n|d=cxX8^)zI}HxxPVf8F^`2+k4?^0daz_Unh38US@^*D1*u( zva`BDNiXwDK{KSXvigUHjbgg6WQVa98c1?0g_IMw+xf`WmspZ9jJ*Hi_8;{^i1ZQ` z7jnH~)`TC}?ObL=#`(ZR`)@_Q(2}n&WK`a62B`nctqnamFrNN?d=|iQ{;S>3NI?F{ zi?GwE-GOF9<)N1&&Jz~?OLa(Bb9wOqu843yFTOb(P;yzt=Tb-z(rvZmtYNo@FdF-d z>|i+@M4>t_GN~F)1{#jG$pjDgm_-DxdrnSL^070X2j5TW6ciL%TJELOaT5{a`}lz zUtEN`-D&DbKlo~7@bkK!ihMY^_UrifgY?XBe2BTa@}C7J1_s_FTwnC~p6D|Kv!oME zouqHjn{V)c(w5%<$d+h7f=L?q|KrX7kV?5p{dU0&Fk$+jM{>gM8Dt=jN>8AzyaFgf zks|wJY$o3CCCxUvOS0wY;%T4$y*6d*zU_qBFiD{L(O0DinRi%P7x66~{@jQ3Mk^`} zmjJ84*ng7xlrgR6n8TH%=DQuE*brjmTi|9Xw1y-+D85u0Lv!00C@Q9dcWFfj31o>#x#Z$X0lQ5f-75L`>x8 zyOnv3(*n{%5~c@#j)S`oaAQ_&qwpTocU@wt3s|@59WBSt&x+pRR~MN-1OZQpRhu`O zC2BB?V5Fqk{XxgLU>D9c;y}0$sdh|+u7R@}$CLdOy;;Wug530IV9N}v&(o{WS!Z8C zK|=)u&8URj=NTu*{h)MMtwuz3Xq8~*5SZ^8C$Qz=jX0!?{!)_+X+6#MYOBvBsUdpI zn&MpDn)VK6!3oNag(n`AHVk^bU-uMIQop|bm;6q&*??F2iKeRmyAMAF9|b3S2WNU? zdneO>iLjC+UF!h`l(289Q{QaAF$_^^8ALGbn5l+_t_&R3RiA{M+P9l@wHm1HD?0n> zcRq5?agU$o@hj*_$y_pp^n66Xv6w`6rbym|=m&;u6=NGIJKT0o!>&RQIbo^q_!Y@I z(Rz~7Oj~}2e2{SX{yN7ngG5_F+8yo&z|_KHAe8_a2GcsAF*iaHveGS}CARuAKtJVT zzCcl#X3|_VdOxK<`>c9xj?{ktfbK;HNBgrL+(@f!a_ec$-UsHVpYQ8w9hUS}q9vRn zE81_7NLG~@tXxFcQ3ko7u;0N~EjmjGA>hX^ZZhEh>4$I_Sn5w3!4TGfzql>A)mVG0 zA+_v1hK`(urRjuG=l?Z}ap9@^oL^})JoM~!_v-9$^qUD+w{kPQJUg@dK48)rDFv83zl zgbLwRf~C);=?GivBzvr)!@-g^JYCbUkKgnzaqvQZ{D6f9T-?O-hraR^>Q?%%i>$(bdvqL>4PiVq2i;9>BsI1>73Py0gyYT2Nn%ciGvOZpVb7yP&25 zqD!if6GCZ{N>08U+4}BFR$JjKdy}xiu)N*vaV%2ialDu)zE?8zM>1^c-}`9SHitLc7oOY4PT z4)ML!%YC)^0am{Z#%shE&svUdj%>V+qP%+RA{bmh(P^?Rfz}k>J(bCNjBoNg(0?Xa z%ZS`igC9g#LI1}T^1p}}yE?nr+x`obmJ~(%AO@7sZ>Y0}deCrWAPiLB?cj|ICFwaa z3v93cMAL6yR%BS|zm1{krB~gYJL6N92gBZ~T(otu3QU6AAq3&lrA~zKkX~!HTbx?O zcg{5|b%G8DtA3V9E+DgrBVo9nqEka@B{e9OwG^wyzI~MxAP1R7vcseXc?pFoc z!S4d}MY%kHty-j1H|OMII!)vZ-SbKBL^}#TK=$W?@^mBCHy)C*n(P}eg(<4f^Kui^ z?q776UIqmZt}BVu6}w}&;Y|_xr9&1F5HOwaHM^iM!|Q=ytM!2kf>e(Hm${}Jjg7N)kQ^#A^5{CBZPQ`UZi4Ws8q z9ig+W(T3VJh)eWx^=i2k7O71*zF9=thC-3xjHCzl%CRrNaI1b0>aun4r@jS!Ischu zgWC#ss*fQ$j%2DFo7}n)#R@K7{OjcNSbDz4=ZJ%77lAR(a1RHy`-O}4ia)_pXKV1o~P0Zot)oTkA$44xzue{|VCe#H!#~DB0f$u;cA&IW^ z`Q6zZ1THi*!dR%!TnpCu4A9?W(=G!#!iU=AEbQ$wiz20VBD+@EM zw7}s0e0~S1NGL#Fd9Mqaz>UGU1?pm}@=Hfeuiibjb!ZS-Ve82H?Te?)C#nKfq;=^K zIpdn{rD~*g(th2)1#L?=b*hkdk{a4XF!VKbE$9$Xs7N^YH2n*cX_Qn9$WpbKces%^ zC@l&DHZ}81fDPqUT@c$vHWYwV6AHT+5t>0{v&rpzBp#FX^8oA!i+V@Sc>2GjmjJIB zb0TV=6oHO_C>iPVK5y^WkR6=9&&EIVzSBn(Sa|A)L znwXWu2#2^#so^+c45gU*!HKH-?57C=$qe9Cfjz^=M+`*ta<{86y73z;`p6ES7%B2Wz6@y18Fx(=Zh|SZoRb}|-{!Ph(S|w~M zhcX4i$0%SYDB*I+#(NaTAT{f9#5v!Ee+{hGfN-l|*buC;Cc~Ww>1*eEaA!a=*STwR zM97LJ^<_4F=+QfyMU2y5swG~l+0(XIY2@bnKAz)VE^jlTiLo|r;9UVi&vwdV!tcfA z^~9Dt&`nwXd{6u6A?>vvuhTf9X5VjoUVkglf{d|}UqzM~>&ZZL39hQ0D%tu{kgE(N zLM4AANr`v%MK$Ql_X3NFif@2;7S531jZ;Fr)p6?+6zcdbsnIg9G>jtsyp*H8Nk)nj zg?n!PQ1gei387dFa7DF639+iuTGvb0RW3|NG4Ejc)LFk~X4!~p1uWp~8EjgMnLXJ% zfq*~OkIwK{Zd(RrSwR_92p`jwH6K?^8?y3qjN9)54fZr+TY7prBz5uKa_gEd*%cr> zna*Vqyv#%66p28dD~bdV$CZ1L3Uy?HKsdzm0{sHE|Dh>xT4Z7_wwz$gIpq4F2L}Kn{6NM`rrru#+tZbkqdM0h zQ?CKzO!glKj4Nu%h8w(9$5svGfq4 zcyR#2OjnQzg)P(%fBhnw%5)#|@jOhrk5k5&Ouqcaz$IdmLci0(0)h?^)gCu42#Ahs zlC`$Mf%%>Y&-%;hUZ|w{PJ?Zvy#vzRZTRV0qhYGQjKXHUY1jFbJ+}{*GZqad!EZu5 zI!Szl=qB=Pf*NoSKp=fDM6Mq-e5j$dHTNlP`l&2lET_A@b9$x}?NLRyUq->>Kcp+% zx23%Qv;7*{T@B@P^YW@LKJrvMY3hZP<7OU2;%kF+6Fzv#{%rnaJl2SwEXTgThP*9+ z>iv1-Z-jIn^n?D$`q3*Vf#n;tsu>)IIN4LKPNA-Xn)-3mte!8MCke1@VG>Tb&9O%Z zB)+f`w?%;;k6~JS9*xx`H8_jN;p8E8@EEjKz|1um)boFPN~`^h;i%po@BH&!;=L)s zmHqQ_?}7Y>da`p-F*LIIH{rcX)s+waA-p%w)Dp74ol(M)_g44C-OfU%J#;lNWi6!F znKpZ$`!e8KBN*@gB(hN2@BI43!&#O)z2j+STRtxng}^BsLqlcMVPu3`xN@}rNBm*U zKnpAqMNE4ajZ1-0Ds~)rW7$!Pb51x@8PGaXB?se0)mobjGWx7223jt(K{sZ!wGflh zAxTcNRuQ@8L$X<^Qnl7ZrX_6AqcwUI64`>S{ddMD8>N;h0Yn+0*V?^=*_4s-SY$#L z*X~*ZjR=O(uJz_yf@xdp5T#;PRvoRbun@;Y>$|HG#+vKl4y9cU2l$`ztH=uuKW)olOS!ytZksYpI?S?lufmci8csxC23?*JVu160 z2gBp1exAZlcLs4@2Kdxu2JHS<)=6+p3Nh8tCip|xzU)N->V$1f6X-*19e}0J;@(k> z0_XERU?Kecx07SOYViSLiMBelDzj}?*QL$t4m55AW-yd@0=jc67wa7J2;B(x1nV1V zU-+jJR)j78SO@-vCis6sNlP!Z-TK3tyML&`e0ndRL1}Lf^+wdxtWs$QiPz0pjay5Vj?SkF(>k!0 z;Um|tykbX?tZbpvD=t5;%}Kry^>p;BKS9nRYY~s@wToP}IaTVCMxLNiYskvMWVI~i z8AYUa$jWI!gRy(DKvzkkS#-!mRUQnC7wv-GSU6@<4lK(lb|bM`Pubyh^}yfkG(jeD zO3kU|whm4Av~Ikv$>YiLN-$f5A zJnD~71W8`Y3ooRX0s~Wp6LD*RTNyVu@k-?y{+Vp3z0@|5V#I==3$(Piw@zrx-T*vj zLp`o?j30Zs-o7++pvpIib&02$$*B0c}jn>B*SWdzfTgi5OF!%a7BXjK2Lx}M$Jxwjvp_B|y-d;*zVUpmw)Xr8ZhSgvg zZR%~MwhC>558@6RG`jSMUv<>;H2sT-Xh~a_Ikm`O&iA>HR9rW%OI`?E6^>=+Hfbh{ zN5ZU$(iSz@NthFaE4Q%+i(Nqu!E?red&IkRM7qHHd{oqi2Bn;YQinlf4~YJ0v+Q}CbB4Ngo6%Ib;y(rlOG7p?A_ zF_RDK&JWms26-rs%5=^T=NtnjDAP!iW@CTYf)qC+Bq+%F@F=FtMt?H|l}HmNb?r+-IxeI&LE@1FBbSk@#LEd7BlO6$#&~TZ{W#Ox`y-IK) z6LHh-2-jxqj(DyfyJ~no1!aD|<~*wfj;~kQ%l(A?&!;I485Xb71p+9v1p0qZV{2++ zX-MxRX~N7%=U`{P*`=zpq`T3|eMWaCrqGE^Qi4(}qSe`j4Wig7a&|4oMN-oBbCs;@ z_V{k!yl!8$G^KG9As_q4IO?u=;+Yk(rHEnHseNn!*7#O-uY;+H_jmcbhI>Dad6<0 z_hpqf+4$|ct+L%cbdc1{hoH}9>6MLh4l{kK6IMX=Ki^?-%{&%*>Gx2P&fYP;Pf~+U zzQ6qqvvq1mybAKm%ln^9Tnn~Cb;cGJxxaSuU^{kULOMAe`St878Z)qD5EoSx9r-p= zy05Y`I(xe-Dd@w)!fIQyg6k++BtP$WdIyicPCQJD=-HT=nbrAzOcyCw>84$+88~;J zO1zrdulLT*ECm6(m+y}HanR7zir$CM#wF$b{evgw3hP6LjwZerqok2tIe@IQv%|sp zFWlPdi_h-%{@PfVlhDtg(nOhyTI)U*mzP@?Oic9pg`%R_vEfPV z9-NqeX-r5-N>b8%cy!LWdwT*92MPZ1@7cURn(`@GTa~Zz%9}DR6T_i*2-@l8XI8QP za=D2Y`>1Jee%v@Z*)i?z?!LIVP*hYrNlhKE)A{K2)7@E>a9_TD)jrKet5>VEZ*~O1 zf2ldK9tYWaE}MXXGpiElR({1{JK&TVL*T!7Sf#nHs}>MsU$ZBW_NNOw=i=ef*^Y@O zMsR!)q^-L0oG+-nO8Y$f0$ap0Bp00po{8PDbmtLgL-d)k^v9Zrjo4tSzx{h&zV;c^ z)AuvTtR*XME?J@xJYj_f?#L)i1QIQE&&~>ZcVD*TRva%UCs#a?o!sJPeOmm{JFG=( zXDs!tRuP#aYXUci%nD{Vv{Fl*!@|ggm7c;f8R&XrQZr$5^_V`W22;evB6p(qV~<1B zDAMvYKw~kK+6sVWWo3nxRdA@zD(1`Rq(2=KJm@3Zskft;G%od#{rtYMGV1P}!qMoSFF0-q8RvaD!W zYzfI0tAhlwCM_+^D3;2?3|V;aRUZyw`evsW9RcB`8%k~Cu>C<#=fm2pv|r17Riz=W zvXe_GEArjn>bnEj2-U~(=lrsbjY}^V2`Q;4w3(3+kgpMJu=`((S4i&k0+%Ju4&aS#C>p-v7-bLQk-PIR; z#*Rhb8a&Em!6i32Q}QV`V(7uie+10dgJ~r#C^mf5v{NNMWd~8Hy`_OLd(0PMVEz!8 zS+y`(8z_$vj?BQF$KTc-zC^WAt7~p+Y6}-09bGy^xO#u*Wcz#hc`u9^&1LnHe+N_B zX+vrRp@-UaJ)7s{&3m%+P6#xn$EqQ0;pq%)76onUoQUC$&_kvOp;;HQL!Hk==NO-b z#jGr!a{&UDpNfLU-j+}S@j+uS${@L}JHKQK82(d`cxG~dNEH?Ny*`U3Dn6pmpoWryp?5V`LYw4}A?)x{1Ax}B zHK&dja;Pu}rVD5Kfz!@dD54Rm*MBBJih#Mj5Y=%F&reQ1J_mWQFj#K5vZ5R!l_0gT zvl9O9JlZfQFwk(WGT4;uHNLMaJs8+y9nXSFA63D>>11Uuqe!7ePr%r$3geLA=t-bm zfWH*VQ_$=4<3r$JEylzJVS^PqVX zkSNs2xP-+Ezids4IV!Zdxp{YY_m7m^^A3K0*5*u`IGS0l@T??qdySi5q%R3SYY6fK zjv5SPq-Z0ii%?R31Th4z9Ll~uw1dM(0KycGy0;E)V?zTicVrG=#awxV7_ne}_xi7K z6*?8f$952BVdv+I1>W9W9vM_PzOU~JbhPYsfJAwjXBOOxzcF{EzIu9dytZ^ng|lIA8g%HQnR8SsyeF6q%OaujUnQV_9)t~&|m zlg=|n!JS$2-}FH(OzDUrZG~ycBQKj%Q*%iWje9G0-o{d*3_kv#e8U48GJKH&J>>bv z4PdCwIv8+#v+9V(#8HD=VrJs+^gjQ6LiNrsXO{cw`>azB=gDB9>hkuQTnBDIK}A*d z_gj2mr(=F5m3SX*h7S7W0I0#AK-(gey02mT?9O|&-b^! z*Y6vH*J6mYN1aK19VI1E%~uf`y_yKJU>s?aOX|#dQiHIsBhs7EcWMVA^~|Qj1OC+3 z<@3Jm75If>uO4tvRKMSNNIkijUK|=;9Al_GJDk`^KV>$)?OB3bgKwV0LR)z#LyUG9 zCZwb_qd`}E=?7fC)%{O5@Sf?dz}V*+fCQ7{2Q`$muJEek+HS&|aAtcSi-u9zUuN9e z+9E70vrb(M(KBF=H`4a?`5bxG$8T>~5aH|e-oy|uS9yN-$pFZ$U`U?TX4DVgntCX? zN&H~1M!$$qsZVZM{T}ck8Fm0IhNd?L&j?>=$|g@eyaAPW$MDYhuJB^bCMc+>5BRT= z2Aa@PJj;b@#W7B=P;LETZxTfXgEF>g&g@pmQUu@A_CzOEr zLxsKo2~z<-OoIfFWG45Ny4ieeS5hxEY`w+BRZL90yxzFVcz?^pB|xvn_43ECCDI~O zW6C*gO(@I*NXSvD%gx%M{Y~OxPGH3p+ zUXB3bu*mQ}R`lS?*A5MLt=|)u=pZq9TvK^;ZmHDhcr0M`$=l<9!a_p=u5*qw_JOVt z0-k2NygX@GJi7}EnJ}afbAX9w4_&?674QY$q(ntQ`I@~Xf`QrgE@nZRr2WW~!HhJYUwXxgELEG%a7NNS;66%<^R;?P{Wv5-LE-t6==qmM=cJzJ|l7 zTX~I82_2s$A`t8|hlYlZCr8|FX=$O?>&Zb-%%omI_TIO^$S8W%2ipeMfPU#OD2V&} zw<&-vkb^KXAeI*xEG+!Qe`^HTRY)0P1SH1*RiNeLdi)3~k!Rz$K5ZS3Lr%}dSguawfdra1=VY_aJCy1=}u4HV^DpPN@# z*UoQmayf9yOB5-oB9GFsKYa=riFaC~D|Kz^Z^8gn5bz;&uKZTNEr5gvX|ji+k0>vH z^Kb3!UdsZa);h}>%s{NDnSk*4Og#gsqbEto{r_}>S{NOk&*3!x0*jS2EhA6 zi^=S}3*lkJFVWUkSJP0f@8C%AgmK+&014Ltw>-LzPEEl^n5g^?qi~1caG4OXwyqjp zly}t}&U}51ukAUl!_FnNIHQgb>Q~tx9sVNC6u?6&|E}S3=WHgXXzuCxV#5a$krEdN zUl0)i5sLm>0|zoz_SnEbW8Ot$L~-~Uj_R~@9uB0F>0)bNyL|)!1=hnan&5%!xvZ}i zVzYs!z-VD+&dn`ED6jVpGc?;`zOe}e^k8e7d0!zEkkA5P>ZL>@C9~)RBmicfq56HX zGTe`2P^6Cl3Q-c^)v>2N%LDJ<8RA=^E~KnHp2-g#IkM=K1y>e$(95HB@7LZ{ny@!a z5rFP9Jd-zPn{?jN(KYcs()_;hU|fZ!-_FA8TM*1HSN(L|%gewRNaQ|(qJWDverX0P_CzMd9N{wx1GaC251^MTISB<9*W2@<^0m(s z6j)734tl}r>JZ6-Coa^{f+IB0g-(MR?3lCaGPDSr-mhQ16X~?B_D;(DTbsGCp1}p& zY#@7zZS0hhhr)fjXU8L8o6&HldJWx6;o;#2%m!=4GkP63g{@L=%+^b+IW9;icMEe@ zA=DgPPH|#Ae8H8~fD6&GpK)o4cU6J5kOJ52+>`ZbNX>&`gFaYcfWr|nuzsN>x6=^E zbl^?s0I!IOiW|`eT4q?h9uYw54+<(z2@^=mafgat`y?*%zMJ{M%(Q7!{q1!cfl#sK zkvc%xbE0rYTihH14~a2vzSqXRCllW+fBe4NC$~4TG31u>b{!CAB?(;f)b)D(XpM>{ zrr+k#WZA1{-`(uSPtY?Gj^MzS@<;~qB=`uv;i4l1`sVHKU4RLr3k8Wl;BSFdq8a#j znVpgIvvv)OjUOcw{1MIH!N5ooL?zZ0D_W>!8!Yay^a2w;K;U{y(kX2kzVpEc&I2f* z24P&U>scY#{k?s+8vPsCTkdr_MPK24URMACkW(W{p2&nvuBo6IWwIumC4Jx~Zt*|y zT(1Z=@`LyAQC9>RePbNW{=gJ(ZobG7iW*Q`bvTXZTWcW@5}2ua@7$aT*aCO|A=U~U z37gx1yvv%jz5>(?*zoe}ECRlE0(D1+1qICq6$~U{0?kC^u;)*7IgURz=Cr8syOW5< z#HzT4!x!?b-`QlAEVyA81;PzGla7^K>b{_<4z;)n)-@6)~52~st5n| zU49IQJbnY1PD+p7cTW&XjB=W8}({uO~iewIgx3fTSXJvq9no1^^h{CJT1Y#e(L z`Ndx3gU|ttGAWn@;wTb=4+t2T8i>aII7Qnlp{sQuotY69Eh#%|MX%pe9NHuK0jB?z&(XzeHR6#-%Rz}eK?7JfQNh3!-4>Ss)J>uV(FuPF ztaA38@&Fc2%L%m|26v>%r~>1F6pSSn%_qUqGgSzJ7m*S5;CbD&6E(tR@Y z6)U#%=@#eu$|(C7Em2e}eNgA2Q7JQCzmh}6&rj5Gf{;jD>k%Y@9+8Ex2+2U5rb(BG ztv*3co~uZlF%^1PI+P9ECrHNwi*}7j;n@&i3#bltD~?wax9^`0SgDW z%T@Qpb0bcUjt=6K(LAF(!7ZT4e^6ahOC)rbDUU)CJaUJgTO#OO8_4Jy+Vyny_4&?D z@IKQ>U2Vg&6A^`&nDECy#T$kcn*_792J$P&=Mg;N?Oz0I6COIFr>ze)BnMHV3BHW&&5(` zdd1r>;yUONMQrG)(;>v|Z*ay36bTbMx%m>Z1y4ERb71W$m)ePxb^svv5ct=pavyc$ zt2FiI^^c*_o_L+TeSCn{vY(2ww=v~Z!ULE9J78XSL7`Ex+x?H5WeV_ZT4pO-hDILc zN^~?d!6QWBF%lAPI=FKld+uMdnm!wxg9S`XOhA7iJW%U9=vsH!nw%-{ zJX_3&hB*Zw5zTe{^?rxhT{^@ZL`a_Ox^_@d5boUzcntY#NV#SJZ_n0;@wFV5H2Bu> zmZAno0JC{Jm#HRM(GbN8#e2h41lp3I*TQDfzFvMOlq21%`sJhSDe|v>V`m-dgn^FC z*rC~NLk=@I56-wA#JWiVxNIv$#Qh*D%6fZq1Cbh+iXz!`AaR##^$a&Xx-&8G@Zt_$ zZ8oqm=!v=-$2UO)WQfzgIrFuH)cKBz@$^E7^lzO_k?8?T2M7B1_IF8^K$27=-6))( zp}cG##ekU^fMww+=#b7grVKw3kH)`bKt;mw?DS<&!s8;(^ajn73^CkR(I_ojH_PF9 za~|EL6%-OOGcg4OK^UR)aLAAV5U=AL*z+s50uGe2@NeSRffx{!w%LyvXBW@`(15=_ zSygp)eIrR=p*Mz)qO{ohoeJ#sZ`T{J&#u4PVU0tEKDP=rG{MAbQ0|r$tcGBufPTY{ ziOQ>!&z4?|wcHT>F;J_>AMc)SXjWH{Z%cQ@+9GGU{oa6VU4T#V0Gz@BOTLLiOzO|9 zy=UzsCg@{eXJe^vYRcBi8OSktqds7Ux3&jF#p(Bxo9QEXD=NMd99(q4mf`5@Tac8+ zW-+=htgf!>ngzu~NrbVNqsfo3;S-O**rqs`^y|X+aebtVdwL+>jD5&uMC47hda<2- zQmBg3S2nnE1t~s?=ab$)$2%S)bK0QUZUHx3al!E|F)|UwS)y7BE!;iY<&X~9lUe2D ziXs>w(w}((%x~-I$n5_>{oHplhv|o;(u_ZQbuQ<43<2mY2i;8x3TNOB^ytqBSn~?7{g{y{ecfUS_17!`i?~GE^a?ry!lE4RWx%v6w#Y+~=e-7ZVT`;hx zGBrj*J0{1hq!A_5%MgXII(s*I$i3`=4DzNtlGU}U)!-2_I*9+A}qAsRuY@tv`=8^!F73wfpyT%1MMlutLWRw)(pi7ZnnD`jr`QKBG8@yQ_r- z5}HdJ`o1a^s^UK!jABH1_pMenxQ#xHe#}J>CYNjj&;S7g1Cpp2_LMS{7_j?1>-p}c1q2- zy<=cp1H#}!dceiFL0%JT7H|W9Wl|nM0ZqnP8Jl(B2voksoeQ9Fr_3dnaW*$UX{w}; zyS(P zUe?90sY$;X-vtGL?-)n(6?*wWNY$rRT<5LF=>@uLMC8UQc^;Y^BDDX z1UQTA?Ei2a+rZe8iu*LC8D+tLDv$A-TsMBF0Rqv|+QOo|=IB$w0I&t>fscd)gp%JM zv79xc&q9?Uj8?JM;eA$BfMma@=<`ib+tl1#-K?&-ODmbqe?6OKdhkX?1;haFcTPwf zf~KNEK(+Ssqox88JZgB>8+_}|iDrfxEGHGpV#(+R#;d#3L7>O@#~egy(vZejY7|>W z$0AA4HHRGaV?k#&QQhM`#)%_+2EYnZ`wgOPxNPke_=K4SeVFkD)WLl;5J zM@5<5*Tb3)BJ7s+*9QCBM7(x-X$ zMLVlq8q{z7jxdWPu&1g|mm4}tNPm1)XZy!PXmY`C6Z;kjpMF>d>1GrZ%k6l0CdQaE zX{igPDT9u~-mmVi4u?;2Q*_H^oHT%!#d0S1huhWlO%Tj(>4?2EA->wryd-qjP=Q|| zd7BUIM`u7ZPj(tA|bz8C;8gA%*{A%`Tt3Nc{ z5ouTGaA3knH*}_Wx+{+UI5V=Y7>E7<6+|`R64$kGaQ5(NKViV+>Ft!d>Y6)f%%Hg! z?l1GZBZBqs<;x%J-n|)bklEbPYjIgyTMJZh)V=pmE_Cj#f+BLTcYkI{@Lm2B7RJqR znl5$>TDr0VAc)JBr{#0}(|AxW0w3XvFPaf8dOZN%k97Bczu8MoO|@;`5bkb7dFusN zg6NtWo4mX{g`lSJVZE6>btEOH)djzNF{mYDU7ZPm^x_7|GesD~t4s?Dq7|qM3JTiY zH8>lm9&sB;%FV^CccVz0{Py0`CG`ISFadqUC*L*GzwrTxmlFyI2k$-ap#Qc>fjM*I z*#lj_MP6}t*#*V{vO$5}EcrbyQnwJlGBgGOhv$b{m%3;X8SbcbcGO#TiTc@vyC^q% zF`_6OEYaoZKYAAX(dB+~c!WQp*dY=*-wH+_g3G4(00*9|uL?)1=W%6=Mt=XfkBo{c zD=#mOOZR8#?OE)O46U_g{76DuOACuYTgizD46-f=umEh5vpgmyW)cpKH2q{K7xO0* zH0z#prPB*YwkqYI;QK>?1ec74SCP|-h?<%hQoh<)!nIt1t$`JI zd`w4~jfJ;v99it06}Ir5P_Sg*DZYPGLU#I1qr}0>Cy3UlTUK8 zp>4WCnPA_Nmn%Un_LflPxRx=*gdXczuq5(t;N135EL!@suK_#81STS;c?TV~B+iga z5l{x4hu#e*|T7!O~V>L6*UeT`DZUbix@atOz zrDDV^8$v8D0Skh}mwM$sHjE?K<7vRc*%XK(*j9TzH8eIbRG1fsgILC6s!YRr7( z&C#=XB95}Nu^~BUrLEgP2K~Yq^?B%hF2QP!t;fyF@@nq%S||N5c-{bq!o|hS*%LQF z>jQ{G!?w5Vefbmh2P8jveqrwDVXEZwW{E<1`{Um414M!U;tGC@p|8nAt-EU=i7fV~ z41924Y=mbWH~~O1Cqx%=*j7Gxzn||bbNJmF&z%Baykc~I;+r=;Xi>6;vCBbs@yhS_ zvOY?nTpr69zM-JRkhR@oYV1Kp26M6v^H+arY2=h*oD9){7i@3cn>J?jc>3poE(1#TF~UfsEQ>w=Gg{bM}Sb zw=@}mg9H{E8{1>eHsE&4845~wH$mLMb6fRTA*pb!1@usr+|#o*DJe-_XVPKgVjY?gGv&dG1dr`F!K%x3E#z4+$IFbt-b zhvaeegDR>h>(l$d+nMLrZr^iLAdu;z$QN=D$jKL<_P+6=GN~yl&8UTXHBcq3YD%pO!l{^v#&wHa zmMA?#Mwcj}_S%k+EDRc%gK)+G~+%8i(x0w6PK-LA>0!Y(8gCaaw*@7KNYsC=?& zhejxCYArENz6yo%^_2%n4@i!43k6fXB(fQQyhA2$XgRTBy%IAWBc*`9ZnD>OH+2K2 zT6cUG^!VlJvUe)+c0CEEE>LF&24sB_$~`G8m9DLv=NNS1-_DM996U3+94-X!TYE)2 zmaQBkHnMi_x(MY`P$H{w&yKIIbCO%##|vlvtXo&M%q0ZQ&%{6;T068%ig-%RrxLFGn)6f>|4ent;|(E!f}{=*=)1`(@XksqAa@@ zqOpOJ)?{}kar~TGe2K(;Fv;A@OPm_hUg7eff3URsTcJg0xg1$C7q2b9NY+FU-&&|v z^;;qX6pSo;Y6>l2>p?oY0m3%FohZWW(d zqjKo+8PIg&W)RP_>Fxgb(3}YP9K7#m)vYgw%$ydR+YKu@{6ns{ntR0>1}mshhnJ|^ z5V}Ekt2NQIV+OG*?#^@^j>i=^jFNIrsY;x-dK;fI8$68~Era7d3AOY=9@M8Rl{8b; zXy7VTx5L&G?h)~a>tLkCv3y-Xk%1mt*&xtkzpxE5+Fr zeGGM6JeU2Bs(P%m+Dm$JLtHb?m_1Rgq&LX;d``zsc1A6pi3tIB8&<%56q8kqS-%0=ayr~AiAe6?H2GS57p;Z?+qU!XsPA{42J^`p5dn9!W@+=z&m zb+hJHxY1SZ%~2W;VtrkKXs3K@Sw^L4Zujn^^2K{~GtpO;#flR~4sUa}>BkX7mnMv* zOaqK$2GVwJ13)^M&@!AOF$zo2g`Zg^d%0`+W{FWqjX8gJT$4W+wLsfVUV~%p!f+Ej zRZMlNB{PSNH^%t|+mY>S>yPe=(u-~dYy(-|^%%)^%kT9EXYQqxymFcUY*$o2Udi1% zYiWV@6HIxZN1(BgZG1LRswU*vNyB{RxDl3EQ*aS2{@Wd24`?oae}wF^B$UlcGRM3MtdPOg(%x$IEG&Xlg6RNN}@fR<@Qcq|kBf zB}#jqxWoL!@7(w6L4!H{M6rf4vhZBb(AL<$l;;y!7)ZsJ+lh4cI%>z3PiFTD=?@Et zpS*5Vq4#($3lorSP8A`7i4w4b&HfV4Xs5WAUaY{t2#<3wgST5UsLdV!?T}GjeBM4m zP?SzA`_a1_Kr8fzQUh|_-34nVzmSkMPgA#;^PLnL_)>i(pmDw1%E2Z5h*#r^EuM3xlfN7 z3E)N8GrAe@~g2Al?{x?_HPo96JB>=hn$t;nb+ItH^y4F z#Ze0p@qTw0@{^$PP2c0%Q`>8V*Sc}v4-Z8U>4P&4v2&Z#Oqz;4F4>5-;MwjAxqK}6 z6!c%HX7LqAWQkwLjZh3aOoTHnSaFzyy%;8lM5hDueSdZM z+f9Fj5ZOdgRh4uKu@?zqO=zOh-iTkumu8Ga&6HwsXG{g(|Ds3LASosM?nFkVC{jw| z7#*OqXvmPJHfNo{#nC4dTuQDfPRji%Jxqq*)RB^j_Uc8n#`}HrwuFrmC_H~wgFG*V zJ1M7;m+lks#U*v`D*};om^3xtsV8Pyj0BWt&e48Jo3CKeH!h{K)=*f;${*BMU~8n; zcTnmPPkgn-7I3PxStH(QnaR6+_co&969h*<*8 z3PyiWIkZ$wpyqaX$3s#Blly|(-AU||`q{u{w`GJU_Htpu-Ai)*{;2R=99T=#i{pzk zou(7h_n%Jl9d5)7uWPz3S+Y(0)N>h1J~j?Qw9Y#Isyme-%y90&=Nad5wg5Duv!XTR zu=i12uD5Ye)2&^r-<@)&EmGJIm>7oZ~q21iYL)ZPR4E{sOZfS3?7DZX8&B^{UaMrYb(bP3^>{!!W`&va^SQ+YOsb7 zOQ|Ss7{PDnivGoDuGicUrfSf_=9 zWUM^%argZ@>vNxj`BFYHKFLDjiCKW zAHBs+5raKrsIDq1q3^sCnbklQBJ%5Z zT1=XOqZa$0qppU?!bFg0S^7Iin}(aKyv{gL4l6os-<&sAmY-3~%z4=?ONLgCY~9cH z&S=tjf`D$Ipg?x0xRjFjZADc1K%58#(-#i~ZzuTtH z(B$|8H(s<9PC8M6q2LF)0@;Q1LcyW31xufM_E5RA1yc?dc5QLWUiCLp8M_%$V+Tj3 zU6ri<7=dzU;&4tHc=m-(t1)jv^NF=d#J=t9K4N*p9OfU!OcqShcG16>mmXpM4N5y%?2r(9zyKQ5T!kmx&4|F^y_Q~n$ zDo|7jKbee_j4-5MAV_7z%;8WX>!CE14oyH-;KWvwQVLPGQ#Ro!^M5bST_HIp@34*+ z>u78q?s@oGU@FIvOua~jk_1#UyWq2`)>zA zzn@-?yy*=t+A$POsJf6&p&c?pWztF;v>FKRzr@Yoe*0y0Viy10EZS~ZPj#k9duHZ2 zLjRrNz^2&M8xL|iW*gLa>&EAPFxR+m`HT?4bBY&9n!Jui+V@AZ=F8y_yakdLg3i5m z+0*Vlwe%^kaz>*K>Li_de&Q(8EEQb#s!l9+CH47EvMGn1 z_ZBI};(AV`%f=Iq|6$T((w1FTkrWvv7z_)r>ha@GyjzFilF}}>?D06Kd z*}SFeyHz^Xh?Ao!iyeN?yU2E>AU=~CzYbnUCy%o)l9|*)A#YjVhc?#*#^*MLWg9pe zXktfi<#%Eq4oZ>8kaN3B9^4csx@=dO4qUb&{p#ee(Fl%t{;W$0YPiN`iJ6YBr@Q+N zuC$Cyn}wp{$Uq{G-OhNv(#gq*k&%&)%~Erpr| z*6%J4YWAYj>Fr<22u0WmF?y|J)cbx3DJD~Te4b@+OX$|_l)cG8(w@ODiApWt$TWvg zQpVHTKq6<7es;H|xxK2Xf-{<|EQ))@+(_&Ccrq-8L{l-ieeH7yS5e#H#v!lw^6qUn zJ09IwF-lprn3BN=5g#++BM9;A-}2>a<}zvucjORr zrvgGyrQZ*c*`Q$$MsSnQztH}ol6)ghBHyLu@`7#1V|9LN$!5A*<}n#z)%41*b}0{$ z@I&WBx^NtO;+vrGq4@pga=SU}l7@@7nUh+%e-KCnQa*cVa?4)8y)(j9o-p*u@!Oi8 zpD#1$s8G$aS*%%Q&ZXD-b${Q|=OT{HK%?67m))AA>qQ#BaQ+|1AgF|k{UE41 z$!@Z=m#XT#Vokk~^5|dnU%ZM}e@2#E>wNopLbG+s5=jT2q>O`-nfl~|J#%qozM=wJ zp?TAwWx1|htpm&+X zys^ORR9?$I7p*Ip=h&;_mA~!@ykwF+%d_qLt`u=-Y+FgZ6T?RK(|Xp+w7pinBDP34 zAu83lo<C-XVBEDw&x3?mc6yegb>rIF}CF1@)Z2nF)v9c}V}{ z?{hp)oRn_#a@Oo}V_q}9sYo|ZRI6Z3_kr&^L>~yn&-+zm`kbrg$neN58>_}&YDGIB zHR4!Z2S~Uv6sr?wCuUaU|KMaZ==xy3T4ZG7J-3qffKYUB1ei zVLDrQ5nB~<(9vFmwllfamM(t)9SLE8 z!RNV4CtGJ1Vj-bj@tEWGXPgs#UuE;cKXuNtGBfXJ4YF642L`7L)O?j|=NT@AhNMKA%6wx_?HjEghjjpV`5 zV*Z?`&sIB=M+SEJpy{g}E?2l&ds9b9E%6RpgP&8J zuoZT9Ckk<2r@b?*^}J@^r?~G7Bks=U@kooAtFfW-)A)=!<@e2Xqc^&sfIjJmdIKd^ z4m|_INpM-`tRa2iT^<%(bHu<4X9y`789L8CN12*@^uH~J$1u$KkT2Y=0Wa*8_Z(Ji zJiJlL1O5D=NI6#V$0idRay++<)73*tn1+iJIiK593-+qiLTK|W!*n^REK^SZXIU*u z)69PJehw|lt8+El5aOWf%o4x2eC8ScvT{T!0+*|ftfHB6{C4kWnVGVmZcipwzu3+} zXL25DxT*fqyH$z*e$AAksbNUM4NJw|m5p!a6V45pU00+xWvq^8j)!Gol(A=DC_S4W z)T?T@pkqMOpnMmDmIhTSmt3=(IW*g(S>qOh-Ut-}A36)E{57OG^-}&rB~#PeE?AM& z1)vdSxsYuGvraivy@4DidhxFJ$aqDwv%iK|S$T?;X|`e{I*f{EajNZ9qehxD@Xo9& zIn}v`%d4%7F>Ra*vxnIITjd)?Rme|O*j@(|`fUC#zD=k)Lh_OO0AbpaN$}tKq(nCB zeMlR@n|pU7IidFgqbhZi$?%pgCalWwtE|Lo_pr=J4E5Y6f{OzaVg?yGdwGwaLXPL8 z@QDI34y2mbL8TQDKc&uX_k~Ny}35NlR-&x)+KdsXnKRyA01k{}*f+Q|CLPKTRCQTsecjTkVJs3GNeg91aeZkc z=nL~IHs~;WB=U5y>rT6ZS$+mb*%o3V)4Z!3GSAATKeKcxp%^HkgdtP~VZp!qtlth*b3Cq$8L5sDagj&G+T3mNRNVI3~%Py`r7=jSzlt zBuhO`=1ye0z#o_rP8-7P(U5TJ%nwk}?DO41C~NkQ{>RACtGExp^`;|+XyL>1KhG2B zeD*g*XC{5i?>DtZ2iNV!ein-4OsQgsS?pt1i4|qWDO~d|Y;X1Zh=@1^dF(Q``ZqBT zy_eJH;_f4nSpws}t5C#ZY+8g2kc80T-6O>|Y$|_S;-U{Du z?@+1(^ggnemAz@8VgIz;{dVBh`R=5*_}Haj&VWua85FfkJQZ)JZbU}Ol9s>?>U07k?G+~L zEUFBL6(nk0$adt~R$A1wg2jBEX}B#*>Wy62(mwS{&~^F`GGtUPp*WncU5=O=v8qg| zOO92$?~)ld`RrY!V@~Qf@M&Q>MEhySCq20mM`I>hf5|$r`;=>Q#Et4{ViX6JcS;?J zsb|K&zQS!?oH3%Ubp~?UMiX{)y+R(pvFE~Nx^h0*CCi#DRx%R%Xvz}M>Q!;TpAs={ zs@-~k-*DvW9bs1*L`-~zYhA&4oxOP4KNhWw9}1^+w|0$2Nc1=w6B8pB#|jmAEe=We z`PvrfvHoO{x>mh28X8)o{RT3-iJ>7X3GeaQknquPx>#QflMc4|NCwX`3)~5lqkJ!Q3le-b#xpbRSJ^?N2&_^y{y%xCpz$RoQ_FKnN>!-SO%fKX8iKlIOT$OrZz6A&RSY62$k%WM{j}yN^~k%+y_54 z;GLHm$)6iOMpxjKv_ZiW?F$+$gd?K9RuR?@iF*bkqhf#elIh+Ts{c} zk+m)&Y2IQv?|EhplTHnaSUzLb0xg`yy?=PJkoU#s%zOezHj!0NGX+*wB86~)AI2rW z=dKxe>LZfIJ-!XanW3y8dFHebtT!@HT6g-2s7BZc!iVnj*+%v0>Tnva6jC3(P-ct2O z2eb60>739x)vAxS7`a4hKi8*+fj?Ze83XkrQ)eSnrprE9*!ZxPf~H?HP_N?3pwSZB zI{iqcMMNY+llia}fEX52z<|McVR)^FoB7~i3j+z{sx^D1UJ1qkoXmQ)U#QiB(iv25{y|_(E4drb?oi}}3BM@B zZV2@><#aP3?A0YnB@cD{!-=S{Z?3S-99k;Y-P##}PjxC8!9%ySQ{ybd37%wYCbpjK zjW1VaP@ul=?stbFGF}2clfk6Gz`(v(7H>E)VVu6y<=4Yoe@@nf?~cWE=daJU%<5ZR z-UPp35b0`fue4sM)2y)uASEv=3nyx9V$z1aOr=dDoA9;pmGgcrBWy_mk4=;L1pPV~ z*&Xn6U*OS1Dt#7OTU(!t`TImwEDdPUP;xPlbwCBKXbMbt7AP; zFi~#Qr*B|@f{e_p-GKP#aQ@fXa?2MC5?)(+7K{!pEiJ|1QV*ab-A5Z6xcK<7si~<+ zNug83Jl38Ptj$%!>v?&!Qc_(J!Vj(|s~uZ|$+tr0dS#yo*hHofy^usKblFv#O|x+J zBqjX~%ltBHZfT)SQ9lq&x1zuFrPg%Xg?Y{YJr3cS`H~l;v@<=CSZ-eOom!dR)nT`* z0+)2v!8`^pO1Rl#I{xCI8i6Y>l!6mT`$Cx27E3F|{Fipla9QqS+#ADrKY$txYW|w( zF*40v6jQ)TBZ<+1kI*Lf#8h}jnks0uxLcikdu4k@#O93m&APu*B3%)wO9@%Q1`wnU zF9vLfud1z+mu-95y04Li3v@&S^-b4{l;F3#TphxY!ODZtW`|V0WG#PwM<38FL@i^o#EC><_N1vdzvPo zB1ahqBPrft1a&y)WC!=s1}#K1qH=%R*OWLf(&XFfky@Ab7ttiyxvSqy zE6tvnWjM)Rc-gEIqWr5te=3hQ;U|*>OkTPpyg&}d6%jxusjySU30*Uu#wLv5&IlHx zDVLN@2`JT_{emQakC&|N(X5BkwOHy8!)CUSKa@K_f)J4q{2<#)2lGcR?iVUL#WEs` zH4-ax4)bp#35`w$Rvz0uAzoDZVV6MG(=~1sKFkoyTmAR`*MlN1j<(K6ui9%Ynpegn z5}hTm$OA|Mznp-8>9o%WqV-sMA;U-a679YfgS%nMeq>gJHiTS>fW^2w*U%FOXW}5| z=b=(RPD8aTShH`2M=Yh}oj#bv^KJAPOu94uF@5`i-2D0qHa;L&Q#3}df7kJ(DvY_WG>-oU_lZ2*?qq%?e0%Pc0@Ura6qk1C4T#uTBw?Kwe&R zv-{ChP6n?FulG|mW_En7#VjQV`?skw^{OwM_x~}y;k9!;!^Ffy61vvQ)02{dLbt;cv*yKiF;l7@ELxm`oc!)`^D~-j!`{sex?KCQ zzFI>*{2 zAqjw;f~dZ6$aC!zp9JwHkY|7Zty*h_-U7?k!6tCfc|-LU?#2S38F*?hp(3dCyvBm7 zt+*Utc3G2&(=35@F~Y(jMan_sftZ0(bZ;hVn&s{~0ml4e^Uv?Jce0q+uiy^N2y0+?eX@X_I-25FvV>6^c1RNG_BA~Z+wBa4Gk&U!#uzl8J1|ShALJQA0@#EFrkX0XzQrKePG@*F$;l?ENlhv5xTW zn0656#EHfVd-D?Qv2%c_7D2>v0UYec<5;_|uGvcYRN+3w=FREX6uooc5=yU#gsaUz zxW0pT#F}34qa1#rLo3Oh0+_8ojT5-N%l(19(O}pT8AiPgaV6L|O9nU;gMJshD!8k| z={MvE2x|nf=piWsa;%By2HEc)b!xg(@I!$}E1i$j7fP}NQ?b7Afm3cr!p;;?1`Aw9 z!Ui1&gc4Ed!F-56m@jfE^}ihU=NdT;WaNVNpvu)HvG2w9X(%i>>k|Vjy=+?Dj=7Q` zeYf=&C^rkK+rcjKOJuQnFa~5*c1}qhViEn)H6r*~R`ay(n2ubv4!f*gNPoP-{kz~* zB4bIVbVSKvzlfhouHm4D)N7tsGssp2^7@5D{;p-r0HmJ9Z3PAGc+N|55>yYw+~_tz zgQlp#X}`ikhc9m?*kLrFzH@kdMb&9<-PSiP>Kz}d$_$H1w~>HdtxB)si>(7ci8L}H zB>t+Ys$voo3%Q)A>^<)@`8_9Qz3pv&zHd6Pdfe<0LuL20F=*94Z3Q9&YPNQTzZ{#0 z|KEnCU%!62f*`63@bj-U*%70n3hOi8Jv>yZclp}eGj~6Zf~*V=hlB8Pa5Q{14MJXL zZqK{o4Z4J^EM&3@?Iv-tf2(KYP;S3+uF&V(t;K9%4<+m4SBv=y&Yhi|uMg*ZJ}>N1 ze#D?ThKx*1c6Ue9|E^1>aoXVF;YCOV_^4r>(xCo}x3E*HDNa$C<{X75!NjBfX1Kb4 z;qE$({2S%dd66D6;Xcfbo6+4*IC_M3+<9>X_LFLQNH`npEflH~N2gGY z%s|h!ZZSUnIpy8~1tcomruHXe1vY$-_?J!Ol%&|j5KX|q`e8KsPQfRVC5NM<<1fB6 zzbXXZ03&cHkX^id3_I*w$>}dxZ@2$(&YIOeF+cv@(S`s*Y@$zYbO+l zQK3`c%}rp0sggu8+lZM= zq)8lyE&1&jeqqk$22v}+3FNwWeig8qnrmPiD1~a)gltF-Z=nOq*CSjmGaIWdEIAcwPx_|LfW={$br~Eu2E%=&1-Cat7nwxpT%$* z0tWGb55-7Pb*?e|JUL^0uO7lb_aoHa=^e&Cfyo&XF$Y@)Db2zgd{6S{#RFjv$*$1*v3#A`3x;f&5=0?iKWXZ9*nV#4RGtuL~_k(J23rO@&7V9oFwPY|9!L_A`$qQP-qG|L!uU)=p^@^)$gv|swN zn{CN@%x?j%J`kcJ7sSZ0xhmf|5<`X%?QuZPHE<2MSJUz9pizV3@dz6v9+ZQu^qwaM z2*-n&fjH$AjY;HzjDL)B<%$JL{6JhLj%*y$za@dj{2G&wfG^PT`)OzEgNz2Gr@O+-3+c-7=%CTk35Nunx{HH;z$9++DG9|H=`Zoo1BPLV@0my~$zOvha)|^v6+&c!h`u zt;=l-eUcT8i{=kb@b zn>l82)hu8R^P2S9(P0OLG|Z7|y9jKj%-dylArLgBEJSw^v=g$t;on_^I~R0@^BWh{ zSEZp$pNWN>0b_2Mcv*y+?oYC1kRkS{rMcBEW#e0KPmv8Yb!mieH6CyGkUga6p9wEs z5vehyQJrdx9kVEP-ym4;E7c>!HT(fqKc8(Ok&^o`;{oMEU5+ceI})JbJ22<=79~d1 zCxFjxezoQ_2_(F)qX)4Bj4^08vRf^jdE{_ydfyhL=)J^#KE)CfWp|y)1i6Cv9Pfy2 z(QhetEIH7YxInW~4w50SqAK-MCA1THes8 zV*+;j2e$7L?6#{F%9o^)MRo^3A5fMHy+lwh66@@-32=w4-F_e+z-Ry~CPVn z;(*ia*i%ah_P=NdivgFXJT-BBtcAK7ld+Z?q{$<8*;fGzMr z;49IBXB=4m2(yO9*E$V%)J?^Wb|XJMPci?9)m!q z3{%0QZneRZbn0BHMAg;Dw0R%!W>(k3z=zs4r)siUP>3LtkANo>l1kJh&&M@uWH*tu z2R0MSo2|rmC0kqt_mef%?hD)Z5WY4~-Ul3&Hs8~Qj2Ap18YLiz!jZvwhP46^5cc^PV(fKW>svIG$Qn1I$N%(ELy1a_^ zwUp5TdPR#3cSR|K|0n^W4X*RJy0i2UDi^EuP<4K$mfmXHnzu=oiNU|pl7(em3A=T` zol(J25J|Q-y}`}**{+?BlrNN4r+a9Dm+)qY0Xt7c^oyBNiP=t|@(V&5A z!+m$L>w{zJxk%{U0ujO-h1p!>42d4cdAOHu*_ng zM@#9|oAPmced(eF3NWb$EHz`&wL1%72I1WVF*kF|8omWy{+nBAR73gZ&Ih&^BHX0$ zJ1o>z=HVYACq$K*5}2>-n@QSYZfG<%Um$G6*J`?;E3pj;;B^Wqf+{tfx=4~)O~ z5`N>^TEphm`p?L#g=!oK&|@e7{=54{_p}OoopKea@akjU>l+rNVP}v9=7CYFf6nj` z>eXpw7`plB!xdSEClbp8fS4^4=pQ~(I&OFe4Y$Q=;Pdh}54*E$FM&n@s9LwBx6Wgd zGtES$n-WzdY$X~{gl9Ea$Old99_psqVW{bMy}b~%M|Mi^kx<$@fN(&k#FfwF`T4zb zack>+(4xwt^og>d?o{$GsCg$7C8RPvcO%saxdV@YVTBZ*Srsf4fK0Rsm^bxn^^u5! z$T8(7{|2fPILiLmNh;L42?r_4gm9d0xGUhj#x>8Yw$M6Z$ z4e(gc+>h`$F%^a+%?;lOTd`bPI{k`JW|Ny5NV9Q-8BnVoBg-a*+G`4->)62S*n|&y z4dhahjMaHM^zkEHJ43{*o_~2{NER9)dS#TUk?6Gi4=YinhAy>+83+wx~%Z zBnCI%FWj=cZ=~|XqXDZz7bit@E1Gc zRG@g%C1JUKImuCrLV!YSj-lW!>cO5e z^ae+0ebkd50hSku>fKEdR;~G(nR6?D4-3pfEknEgpE*K0h>~VX2&oXG+kuc?Fz1mosun8Zq9DQVWJk% z-)IdVV8i|y2xTK4l@T0aUldIsgn!+$hd4kpVa}b4ou&AVZ{LBKSIb2$OJQ`wWb{xj zyKX;G#=g95^`?cBZ#`kT9h6?#+=^2-{(fVsY_v4YBOP`$;b6GdIV9>xy;RGoq8}{7 zO1|EX@<~g&Alw7#XAJUzdso3gokId*NMUSIRQ^8m%ssKCbHNJwv*lR4Jr3W3j=*Nc z>f8C}Q*mN~nNy{i>1W9ISg&qP$Ya7U(ux}B? z{%`@3SiJyKnbYi zw=ZsfU+sF=?(n?N^!s@By6kwg`+Tqy&i2DEUL^(soM1t~IM(Vdlw*kmh`i2fA|l{6 za7;OICwy;)sepH1fIaSXIDrVkJEi$iyUVet^Y7O3n(ym5a6-?0I~ZsSC5~Tgztx=M zv*B^rU_QMwBg2Gf2|oF6k0CU%xBc7G^%t%O^>xp^e>gDL`+a_piiYSo^kbf{H0&<= z2Tf*Et*HmSMt7I~yJK9UTttH643$Y(56K%-H2kR!*G7X#YDSrUiVp1^dpJ+Mq$qCNqpS@|o9WsL1gQ{&pb6;fkLpG_^dc)hBbo5lf zpjq~@#WVnea$EyzNh;LJ7SZ^wBD^j%#`kkVPacAD?sa()jIU(s5~H!}tx(%bkZ5Qs z?3%5w1W+de_G+-*`bvE6H~ecIe4midWqWU#F5tLZWy8bHcS~ORpwjbUze&%>%JB$U zg|E7@%(jzhDyd~u1{yi*?rfB7P+prrQo9?bfl0ib zO)O3-$9XY=e+din^r|=6#o4-Fu$Z~|F=A!sL9)gVK}$Sz*Zhdh!UwIqP1Ts?Y4P-Y zUIho?a}_3)M>&4XwRI2TyK`3n%z8g5=@FKC^diP*F+5nkoH)N-i$hZ1^wgdl+{|if ztv}cW%mu8w>e?c3p#i@^Q{o-@rpxJS>$(bb6*Ge1^Lzhnc7Ywy$GyJT)1><<6`n1W z9zKu9M#R%M{uM6BQ3rxBb5LGkg>DfN0wEXNFmly{Zxq@>aBh6pgu_ij?2(z{28f2Z z5L&ndcB?$L%^___lv6J-667zWXTW!8J^URkf^9Bj!KK4shgqx+0T0`&dXH1aoKWIz zcX^`iBWzp#kt042LtTN}8*MXAJ2%T+HPVmmji|0N>@aH$25dd=@s3#j+X=7hMGEjD zWL^(w-Zd?V8(xP2CEpr@W@Ok$p{n-#3veH(Aec~m<`-k3{S zY}@*P5Bofq&Aj(jx0dU2C&rU|cW-!!W{BRMs92nJZ@<)9dvKIOyQe2yrVc^IAZqbc z-?jX!TsB8~Y&_TOUwXH1C;Za*-xF6H+A>_%SnkU?eP?#H`TM6#BcavSa(<#qp$#$C zjoq*rUTh1oMkSM$U{`0RqKxkdw)OdWYtXNm1j}5i%Ol%*8-`vDQZZZj&V*FKkQjOp znvC*290^f8#a+tvnLUv`EU{IEv@Mu2@MU~NHr~tGefYg^Uiahqy)XI^!Z`PZAQs7I z#ke^HOEEUdNO&v= zUxJSbFW~3f39;2a--em$9{@_rCtZ`)#6tOuF!3B-R$)F!T{Y)Z;$LnplIke z5CHvra&n@WCw_mrX!jMsztG^{0P!c85&F&m9*0@u~apwn+p03go?irO>l{9+sK)w@v*w z0SQ9=@hL)W3Ft?;)Ud#0LUXUUzS=@l>0y>TCn%?6v2X5&qVk=5PCO>>+(EA8qG%CS z<}hnq)k^@?-70xyGf1CE+r7V`NCa9;DwIcY2er2?R)%Ef#2kf6Q)?z2`*~c{uFVMqUR$+65q5z?BPlx05lYO zdV&Ww2i0X%b=9pBB6r=1<20CDuaq9@`MloPy6L-~+W;SRTxNZMqR_RqosM<}iUM9j z#HkA25V_uNB-w7CHzgqHn={pwxvo#m$B%+HXWcmtZ7e9YfCZ*YPjwPMP|@sl;0+X0 zz{;X{2X_6jxO$Z66@@!SsvBU)7K)P~b$_Q}9_TZn(@x5blihggZIm9P8yEupiU6a7 zE}uT8ullhUX&15@l}uQPyvEBs)H#Xl^As)|xRHwDNB+~xyJ2Xwj|wWpZCF)+cS_G1 z#^};af1j8|iYIgm;{c9?T7*z1J@~gMx1lDKV+3`7YlHw9)dW&}_*q?UGGBPW-}f%J z!ZGB09{X9IQE4z_Bw1d6_aa=1U-vJjW2p)~Ny=*{=%$anBr zw`$3d+nO?5ki*^jyO}HTizbq)Y@eknH{eS-VOU{Os-op z08j{0{Nw3X)FVsK8;a9=R<{d|i;DCNTYp(#YR6uNOg?%Rts*z{zHjaRSSU!pi|}H< zcP+SYON651lXNX#*KIV=5HjQ#-b|ElEj&1Q+&D)65c_~J6{ZfIB$#^L`!R14i3k@E zGK`oa`l&7+%-1AialxzS#wSGicN6!Q>YlBR!!*!=HQt0))5MyOG{9G6zkC6%apyxh z2N6Y06QA{7ddDLfHod6Jq1n-P79F29iF9$le7)mf@w?yc^w>gMJx#r%IHMvbPjG`tkP{9njR{BVA3H zDF|uNt{EBGFUI$NF8e~jw&QL_CY{S}W2+wO-)?ItB!VISUv7SMw6wPE*8_fUgMRIM z2{K^Nh~L?!X0tx-iiDnr<(~wm-kcWvR89u)w_QFqxSga0@rSPe7u#9Pr?LgS9UB?{6607pBUCLA)ZMHaGcDc<)4FzrcngQP$#hPUhHzG4PnamYt|-l0z@tj zL-S{IKH>Y(kFIwIp=XHQod#nF<$3m#{ypZB7c2NewmdGQ_u)=KH6^A<02EZKTQm_S z=aeQ|2&b3nO2$5#_Ht5#l0IzCBR>XUn1eftWpmsyIj-K*R4jMJv2I$CZQ{Q?Bmwew zu(C&!NZ~8#*HLeF%E+jYJ-kL#UT{S?6+3sgMiK}UP3iGjEl0Mt6F1VhttM_}p157! zP%mS_%{*V%TXH_5Ylvm_>oc>=U{=6--OL!_SM)qi*I|k2!@HKy{AA9$Z=zQVCv*k8 zG^J?Xd&K~l8fR#pk0Jmx9gX%%UTz!>i=m8Jtn!YA!(9MMK()W%c$a_&!TdXP zthL8+L)Y1(l0t!S964c=%&05myXa;nukoB0DTeHBZXn-I6Kv6#CNAsy7HT3`nRa7N z$YV9n*7W!c;cg)VoIiL@R@2Yh(iuWFY`Z^Ax%AAEl$JfG-Atje>8NjrFI3|?142qi zLlNU)r>Jl{LcyX6S1mSwMH+x+-?6mo{CY|whU?TIICY4^Bgb_i#rVc7!uIf922{he z0WlIAefdFsD9NgKZKRkEv!1I*j|YfEf)0E6%Rsboty&tC&F-i^O-~qOu{dO`UGc4n zCwJM@c>Bw_=NG=haeZy^stN;xZI$eGyO?yM3$xyJ71;M;FNJ>OH|AbuE8&npEWtqR zZCi-FaBQ9&UV@8~fI!0w^DJ)q=@r_7@E!$?`E9J7U#N+Cer8(!*b78;OGjijq0o9n zvh6&sFS#75r^jxHeIh2@X2B#IxP#pf=ZwxLQKY0b>M&)}sg9ihd%^hq)SG2;O|>IL zz-@i@jIE1#;Y2y9r&!q4Kvc**=H3kCc`m>0i}q}1wooQE0w(*@Aanq#Z*X085Uv$6 zCrb1E>aN)CppglY=$(cWD9|Zo-9MfWH@z+eK1y|{p^*sOzo1%P@OjVk{v;ZQW#?+U zFU$AkcyBoIvh}pGy9m1Ts+ah(sQt5uVrseDClLD!x5r%;b-CQHwrG?KA&TFE?)s30 z+P^aR=Xz+u=ie^@IqVLM|7QDlCb_37>+|h!8>uKipWH8kjLFaSKl$(jH}HtJtJ_Uz zez5)d@LOR0ZSwu((aV${;#fLCS(t3NaPIfilnyFs>4%ZDM+!ZTC( zTLfolV1wp6J#HEY4Z*7{M_M!B*$27%Lnh-XYOtFlb4{75zWf%ngS!DWQ}s9ri&D)i zCH|Z>0FtYf!o}|a|5|aW`F%+z%;U>Y!MSoT;l&qA-$p}#q0TQH21^Di>H#jE{9wbA zGH^KWD{^3xeedKiM>W;%(C7nXM*U*Y4JEO106TXyQJ7|}nhf3aoiVXPxnt9cmc+Gi z3iyzHcqU#12^Gp)BcMaJQWt{@Qkf5(JaJC*6(5AteZxU zp56xUbxiGhr9@xZB7C#lBqnjGv9Zg5$b9HV7!1s>JH#ZG)<=oyJ zsyc${V(hOvL$6TrOVOsObL&$M&%bXbd%0(`Q)7E>kd%6Bxo9G;a0^LYK%Ba`nn=LO zrvn(YRZg$zyMD=+c_ghNgqgA!zr#+%{W@f1@+sbo{Kqe`ubz)y+xJEKc-ID5pz3a2 zpN-Hja;cp``fH#|g;af{Tih998d3cTIA=E{l4<9VW(mDwJf8JgP?Y$?f0Sv-)8D?oZ_moONZsih-f23>pr_ zb9t?dys&s2&V_Siw-~+G${bs{iDJ1 zFUUkK?z$X89+Tjj;mFzEg-ro(e34**q0CW#O-io>^1egwsE2N8fu+*M_Ldqjxnm$3 zZVb>X$I*eP9tD!%LFA;l7SlWXej}k-Ehc&44MMIPVZ$Z_S9t6%Ug4M}nk06fUS^0O z!A3B5x3&z(+<{p%Iy}APFHjtPCHAq^?a!3fbxc2Aep&St^wp~4Y!aHX+fI}A6r}Yz zgrZL))rP7L3p0$jUvGUhCj}y%=SjqN-Y;mIWw@x3m<2*1z8vMnK8ab#X8&8cjZFm< zH_pH#Jr4UxG9K;F%mdkh4n(IlC17lEv;E_c|YM!91*5i_pJq7lw1cB zuDG6Xnw|taRZ6v^uh2l7+F}PtG4hrY6pMfnekA+8-!G0PGqj!6j>_>K=X+hW6qT03 z+3A1S0%ozU-LzRdj7^eEh+N`@h!CW&#acG3+zXe9LhJ-N@t>54KR4c zy3EDJu?5y}>5`QDdhBU*f(U7c7wE?XjDy`&2{8*5W7v&TTVM8BZZ-e@#fPj2&%mG< z-iiBu)UFTc9RvKN5IoVGA6Du9Jzxl@of(O;p&7}r@tu-Q0ph4!_V@BQ#1A3NilScg zGO>&)t>t&sZ*<@V(4shXf4;&o)>SW7?3lS}$to#?v&cAMy=J#&RJ4KInsdmKCQ|TH zguzb9W#>_T#xxn|SH;pz>|myiJZdJk#L{~UX=99jmAWXxynmWMeuY`@*&n;7R+J4WA34c-2}A z1TPB&h%t4?QRzwh(b}(H9zlVhaZnKpb!LFqjw2;30(?xxc{a^Pd^Ik_ep}i$)T}z&iDm z=t(7~`0M*AJ2NfYkF22{1pi-N@(r6NR*!Q)6;mOEv3P50!^N$_CNSHi)B_?1g$@ve zCCpje;u_tY*vZ;mRx%{o=xEZPO6CU$2nQ4#p>d@+G(z8EO>heCCBAv|Cq@*qVZg?rzI!y2Zo0MNOw!KS#kr5 zSseu#Y~5R+-7rzuLi~&Uh$eLKIA1@RvEn&(Q)v8WxCK!`YD3z+`^Y8;MjUuqX=n?4 zz*iJ9wWTde8gU<=DZ#Oi2&JP?5@p=F!V{lH#!K3%2U$TeL&7|$m3yYS0@*u_k{;EV zX)&FTOP5vJjxXI&EID36VyBb6HreUx3i7oEujdu5NA?i6aEfF==0|n z>g}-;{2=Urwf$oA6kXS4dEcAFe>Gokb!-rBh_?pWTc?& z%E2{g4rfv&=pQ5{m4kM~CmO*?NT>(KTnuMgK*xu8nru1gT*QU*q#Oqux5J|sphWf!w5s;c(u%<05K zo0S{XJQwVYV$5+!Y(kxC*TTdBEFI??7zvV|yN$!oFQ)2<6?{38qRPNnT3V_x5_;ZC z?3Dm>1oIIYaYH?_pX}@E(&+N_@x_)n5Q6d;mYGie*MnHVeD3OOq_c(bewI!}VJ1pQ zi_;Nm@8tJBLwStWruaJM{<4hvj##2)ER8~jV2ER?c$ zkh%PQpKJYd$%z!v%kUDlwh(%F1U(i5gtsTR-J24|T1=K%)TKqIe%juU;lL{Zjy%fC zT<=>SN`t97;xTC~+9qt*(@yRWp}j%_j4lN@qFh0^eB;U-g>j43E9}2I99{;Hp}Eq6 z_fjBym`UXo#XdHBUKLDj9|vv38C|t&{_)J^_C>GkqP}S8-4k)cM@THE81GO@MgekO z5F(X)jbKh;Xv{%chn=ierq*k_Q>+FK1Ohu@-=?c!^8ZS-PBotZd_>m!g)N0+`?kE6 zihWBlkKD3c%o8VdB{*%Lnb)A?|I|2}49D0j!qbB9?vgBF#0JTg!mGdKZ?pa?0Z~5# zgmz`=`#$f7p_1WoSl<0xZT$XnzglO8+~q-bJIF~jMV0T34MdtnH7)h~{3y!wJh@-g zCmv$&iX_X5Z^!T(hILlNAe(9hXK4AenXPlhWF%NNEae#GER7|bP=9B2k#tAK4@EP} zij{z6^i>$oS<>=|yCbuRAD*l_$*ku66Qb~H>ilz;{e=vCu3e8c9(1EG63ntGo zjrEO|zXS0PP^2rDChFrb1NM+sf!R$w?dVW{r&~R24p=+?_V;^{Uq%ZXbOV~1EHH7e z?iz!U*IvQxb~pq9L7ZpPqnSOMm_-riMQRZ;&YZ93>?Q9Ysb9^+mee@uZ}QzyKDoOR z`UXngdrV{ug#gHs35&G|o52e3W9PK0>*L(cPg_fCr9~#C>-Dwka<1jJRJE-0<)n=F zk=5x-3ep)6X|J(g?Z{*#$!Suc=|?)}W#_ByQI*rqO+v~~IvyZn#7~H{uJ5VCPD$=@ zIt7-wBBnD@Wyj-=zho*ChUwXA$Q^%0%~!z=d}k!zP>f^m7|vk^p9l0cfTAZ&G$dyF z3bOmDwxAq&5v4cf(hrw_43A{Mdz<^05Z@4>+;#U@Vh@zVej?P2IQ2PlA}hN4f+rYa ze9)1=A}v=eY}zMFp3t>^*AD6Ac0W@=BPbxu|H0Yrl1-f=nuc`lCIj<66K0cXGMl%Z z0A-u363Dyb!Rb`AdlfXjqm(=zwa#f{fhj z(9&84lJP6(w z4*R~poG;aog_#T*gog?OdQsP@m)S^= z>>_{b69Y+}Gihh|Ln1MMnNbK{r<3kOHf8tbVN7k!B^{XD;p#cQvQ^{C zCO)rENH4`|o0xjN8>$ZHURMP+aBiuNspovqtaS%6RRs>o^Q{k$d4UEa5{8e%yaZ$E z=WpYr3^2)C6=%wtd*;6qY`US!8*Wd%+UskQ1mvf`o*Mg(N@lUj8A_IkQ$QJFTbX#T z)jii}TDJF;wJl=__&%O{0-^uiX0A5uOO)d+`J$fJhx4-AZA8ZZt9fd?XebgQ5r4B{ zo_GeI+rMX?zM4?kHzp67+W$-$7fd&`GN^tUDt3?G(q1p=`+hAU75?aZ0_uJa%#Y** zCLhNHqJ6FzWE|JxJdBV-+<-W8Qd1I4L&y$4vJ1fDFA460Dtb4(HB&RgZ`jN?zxka? zM8|UR##Lb4;}D{rE4vLu8{b1nB=;4I;kvhm0SkrXd~kf0NE-ur*%Ia2X1TcVeQUD& zvzWQ>p2#G;>+JKWSW4+*c9xPtsD6fGXn&FbgO3oW@)nXB%@HC+n_M%rdPCH1BWwpr z=w;Z$V)5k~6`pFgjM767W=Ytn@v9eF~v!)U|11xte4r+gjY$d;VDPWEi z*F!B+%ME=H>-=M?rSRn$gXH;?v7y*00@xNquc!Wgl^D3<(o(bOZ)i2*Ix52h!vMb| zMwKxvGVcy&z=;8KR~75iXwVbDsN4FszWJ`3tv{N=^m+3|L{_iQU0-b{;LCx{1@HT@ z^}Vlppc%9qtryCdhvLa%33!yA&L`XV3YPwxgLqQ`7gR`>C2R92?3lrMPHSa;&5WFh z#xaIVuj-W+2r#znOgTu)ncl#H>SN~b$qnANulIeS!i2WwLfK_EB?DukS#}0g#WMZG zFh|8VS#SYgo;w92?K?^PKVp{wcal?Ug1?pVO=#D5iIt**VUQz3IU!?Yt~IS0Tjk!| zR@w6*!s3lE#%g{UhXqVN?GR$bCY;M_8)RGI&J5Tr9LV=nZHajY87(b^9O{noP5q?= zu7Od+#98!bn_KDmw!L6j@b{NO zYlbqAd*aF^5A)-Y3gkOnd&^xl^>y#f*ITtgH@u6ateQJjgy}oX2P-nXiY1Tc)`y{o zjJ78;WjUeuyKr=>h35lA{r_0#5&h`jqpX;im`xj=r!+->X3Es7DzuwIGF622Zj_w= z3(II=(VNaw?51CT68=Rz#*r8MCUZ@kw-L=0ImICbn^gVxZpSof-%C-O>e5V(ltq_fnGis5Kgz5hx`Zb zF?xiF`0I?|`S0}sJOQwRChuCHu@yrM5mlMA!>c<4bV^`<@yi%VSQVJJW`FNmCaIff zkiL@;cgvuB-?n{Uj{nPN$PZv736~m0y7Ra}VCU;yG4@r(bfPq>_mn0hVKKy6JN80T zIht!6f!9KUjB6%3AFPkCA`j<7pNAd49_Z%nf1ithaq?DyCA38D)-X1QxiqZvN64pL zffE5Xkfzy$N8Cs3rRWrAVcaB4l?1_~B$9+#KqO;SuCzPd?X{2%OlqjMl2vP)F@BUk z4lodR^0>Y*X}Rzgo$pU}@p_8RZ!mvPzHrlVM)G!y7>8^q;2@4BDlR{COA zD=RBMrt1;dJ6M`c`D{@h?4t!|3>|-5^No#@r;uuGi<8kdBfmD z22h-w4`$1-elNiMAB~*^2zxworkebw>vz>r#+`Scb=Kbm!81JILoJo1REx0$euX1a zd~={fy_8}$>&HTt*uazo8#{s|cYyJu+kMQMNrxU3k(33u?R*toLyo(=ljCK99%JNI zgyO;%pEc|UV%NJC3t++3?JM+W1{Igc)ljemiD9Emp9KX+T16X~^h_b$!HW*C^FQV? za;^P-M$6GCYD?i=Gn9voD%Tb(=-H!Ln6dzwM%i1y!gW8Hgae<5+L!#yr?(_l1olKD2nB+0)T4o4U~aTO7>t-BXX44@)t!oZ9zGu=g_*4E@@C_@ zaW2^I@d~7IQqx<1{}^psf4>+rd7Azs?MrESxLtb=ZN|8<^c=#6&u-&mH93e~QIQbp z@z=x&u|ZlN+nneaT@CZj(CYl2Y^LW+mp>nxg;VdY z+YoA52+P(smc?Abs+7a?FiQ>G?r^#a0xhFW86{(ENR>jjg=jq*3fox;dOt>&8V z=FJeROl|uMkANH%6i+v~pcu16s#YqY^4qj1YtMyA0n-sYOV6J-=X-}J_{0ToZKdPn z>80$c{e4A~(2`wGsbhExes5XtG9;8z@ts!ieB3OE;kKXUI*U4dHqsziqVgEWY4g#X z$)B6N@9D*=`we&m3FIo@z$rFb&1cJ`*b||&sJy3-MJjf46G@9}B(QPx9hxsR^q8x) z8Yr-yr&_8u<{hT54IxpF!e>+dEPjlzr8RjSM_=QYK3=bKrhbG>+SgRku0J2m2KsFj zi=D=BMR~f#c>h9iv%-oFcp4I;;=*xQzT~g3bsj_q56O6|%i^+_F`g?{^ds5-pMFHM zTY_4&TO27qb4Nsx1(N=q8rfm`hqOvuKyk+WvtZ6c5t#u=m*gP(Apu6R5vrT4kvQK% zjtyRB^P-_wIlfc-8x5JzKNW6+7(*G=mn6PBbsUjqbjS(}HiP}w}fi?&gjoZ}^hHemsJsx^*A#koH^#Rh-nDK@@u)kbq9n$l-<+P@bv z5IYt#P%0E(l#vY&m_hW($`JNwbkuzdDCQvg`I_uWkF>@qVlDij7pZ6?+T3l@Yh)90 zIIx4x;SxcC3nJ)alJQ%P=yZKs-s-3X4zYZ_NfHM#J+<1GEsguvwTR(gla^bB=a|5{ zlw=)G)mUZxChxuB6m^@87LQe+yV~b68s0XTpV!Tr)2gn6>Fq%e((?Rt4Y7=zleW=3rA3#>zjTfvlcbm~*3{{qQt#BX| zA)o6XF1wAxEWgZ28NG;>J&0E72&yR#uV@IBmUmh?sx7#=g-~4XK1EM9PX^Hh=g7$= zJ3jg3j<_0TX>_W6=BfxfD~0;fpd7p$7Rschn|knr?;C8)hH@{To}3jSXdpQKaEPuV z%G_$MaFTzNcM4H8U2x}!m)afiPyeXaAJ7<=^Nt4yFU4NLo^YvEq@uU^kN35RpPpsF zNv?6jTqFVaYLM!@bJ6!^aBs|PP){693{d^WX9BUuh;o!i`KAuq+H_I#EX(Z6OEGcm zTcjDWoS5r$#vPoo#KlXPLyY-vtiU_#Qhi6hn}yK%^X?27y=Q?P2Ix6IafPM2Dwl&1 zSPG7oCX5d1JFkHw@vy9v}f_btrW5J&QzI4;KoI;mSDnVoXA zU%jk)j%(DDPT3kRX&~3UuVd?y?*Pk*(>VRk5u5IL3#Nl@RvNXN?F~t^H*deZ9+0i? zb>09DjdTTnd2>@lLpnacun#%@?RAH7|pfYjw5F{mUdz@@!XT^f;MeYgE$!(Oo*m8(bD4BC^T|u_SO11YYasT zjzn5A$H6M4K2;h8zEuWerF{WCw-07$ruZJLtAKbuCLX9Lj48n#6a7gdIHujc(}2J4 zHHM-b6aX{$v0^R1R+Co*S&r@rlFGv8zLV!pD=?)1nS6YuK!sjj&OcV;c)wmloo9=s zM{L8HmZ`*i9cN04UOZQMR8?^*3^q^DbH#W!VT zLB`ZKA_Zl(n7sJl`4$ZILtEcbHFkMadZcj{xv&}U|MB!rfswFV(`anlwrx#p+cqY) zHL-0>Y}>YNYhv>x@3;Tc*SYD=vuah4jR~Lo3A~oz{jiYFJ5||DHO7qpVKxueBQ}48j9Z^1qFrd z6n=>xA$YM`ujh3$csgHz8ZnyKeUZ})C-%28R;lkHyuYxVBEqkuy(B|oBtHyrb^~2INi&Wg_ zW)S@w$^Na@A0x~0%y#>J0lL2^p?29pn)C*M4#(g>9i%pw{PZ%K&*aX_@ppE%NP3j#v>s-z9MK93Ti5`zPPpujs5YX(jP zo3O|tQJf(XF|B&*O?A-twv&Z{30NafI)gxp1dB`K*d;xyY65*m#`IhGMf&GFkkqsW zMrGNdnF8QR?j}%W)b!TVEbSvu`N1cL69hJ>;sJC9SwBL+Eo-5T_-AnwdSj%AMB?F8 zTu@>rEPP#5;oPWL3lwrBnC5u}oNfbeTK`6IJN(GhhQZ8P>*#f#AdsRGG16j~+ zS>9X^w`?-WB%in-Q{AlwfE7@VI+N+*-%6y+vfhGL1Yr!!L*>E< z5gD|>hlkA?Qg8TTbR4`m<_>TQehrO}hrZ|B6l9sccDyhAGzQ#LDDdWj2shiyw>E zT{l9inzvJ99X&KkORu%r`t4Vz*FipXsJ|J#x=O9Jkc`FT@X;S(V!T_EKJ*mO*dZ`g zUZb0#b-WYLZ+Y~uZR7bU^@)GYcjw0$>xXoDHo`tAFDo;h%6$2`&Xvk&v{a!{fA>0E zofl*duE}G!)$wxy-AipGQpgO!>Khyb!C-+BrgHX1q!|BeNw<&rjkA81!*kCw zZ{I&~^R0i;QY0VmM3J%ry|E6i`0aJZo;jb@tTwYZ7k6~-rZZa9?2GTOD`wp}BC!bv0ylQvEDgBswc!?&@80%%or-W;mpu!O#bg7pL; zQd~g_lg0`Hjir%gNSMKS{riN#LwT0S{VQ~XZjyUm4 z)2haav*=1zhI9r}0Go95_#S|9fNa79j}U~!j2MksrhXts`z^m#B7tIH5`AFonMt$g z2mlB?Cg{|(Ad3i{VdW>Q?jWw`TCX8gn7xz#akR8qKXoLTL@cY^p^tY6CL4G}{iS1q zNIbq@lU1((0O(-Bf=XTgO$el4;6V>{?yTeF9E1g<&TrEF_@v`RXNW~t8%ql$M zPPmH~Z-NqoO%2dN+agMU*<|5op%njuY%DsNNwt9YN}H73nPa#7GSmeeQc1Ty*u9*k z?;^jiu9P~^fgn+0aMQtQ>(M^7Da&{Mu623NYc)D_iwhp@{Wxgn`?mgt=nvQDpX~i| zn38jRD4h6n)p5!fi@_i6-s5_w7Z#h%db90ep;YOvnB!Kl;eS?vPPqkAsY1Q%oQpQv zhbSe-iwT&dpk(EKTjDE zNZGr94zit?v5Z8znTjr?i#jlXnEZ>pbgsS>GDF)mHUurCYz|jX{pOxg{~J8a4hmP< zHy-7yhxODOnj{~88C2y?ih_o*7c=_P10+x=>z_6tu@y9RK$hu3|6=K`s8gto*hx*@ z#tn7m4;M%$rs)@5^l%a5d_=F2==U2FXHar7k57x3(m7T66B7dod#r*Rs1%`MDLz5Z z`tL@0NGucHtLFD+Dkbz04K<{VE96O?a!_4~bc_g1r9$?@gl606Qg+NT|6AYPnjWXq z{X!#Zqe?rRx68Y)bZ%P7NAj3m%h92n62;MXq{w95q2dR8KCTn|T$YTWYdeW#o^tz6 zNBp19`&_|y|GyALKB4vG2)z+%jyGq$lPH@XFhU`h;ZG)uYmR;NYEOsbm3Y8}vs`RD z5TD;+YMzA%K2RBTM0v$zX5np<@4A}_P(i&_+Ir$)vjhD~Y-WaP?7KsNg8b7XPc%-3 zQi;fCnx^uq3ID1L&b;pYr|^zZH4C)6z&ci-Tl> zd6SX(5^Jv9_q@#!0UtJ19T8By_*2sQgSIntxjDjWE3 zmM%YlqdJKynshx~{I#0$qQCZ3nbMi~BV2EJK58F_NIf)qY&~k0EpHIO`1(bO#q&th zL{+_FnAQ>(9hW=I(C1RAt3WVhK0iw6vdZNtzngM>a@G`A@dvS7SBIO9^EVY=^yGWF zxcg%C7d044PT{v7l?|lcR-~$^sOY%rdOKUL_FcDWPD)Jl`gmC0=BJ`}=Op|uK%T2Z zI)>=@`Vi+reCbpbqmKwZZIN1+9>lYm9decZn3X2HAkk|H*OI5&*oFkS@AE__>l9g! zA;BvBj@KFKFF9KDRbGU=_vPH5TApfOzW+Z{lv6EqDQ`68@|-E{T5!MO`EE4CW`bBS ziu39Tc)dpCdtXXmF%C+t7f6jq{y}A!J`Wy%T?^;3ZSJvNTgtLBz}Sf(gcsYIBKEY& zmeRt(0bze2Eu93|MfCau;xW_aYStilw7v?l?OUdv_Z#4H);=Aa;RFz*s?xzCTtLIZS8_Uw?oLXmbX z29cTNIxvLBm;miOvG!p_(I~(RJG-~BoApN&Yp8_Ksn`krMMUZck_C~A<&42Ga+$y0 z@vr9_Hm{~MyYh`aLGkKUgF6aaldU(u9@`zcb$o{{WY#x6^}8+ZmJ=5);6Jp8{6BBM zQU#9xu8cL`Py6eAy)GAs#n`SjoPwfjGnh`K%JF}&kk7bN_WiH+j=$bU&Xff8iq!To zRq25HnaHw_q|0VOXl7;E;7=gRB`nSHX*R}s9OgXiTp=cbhm~y0Iq9irZox4I?SJ`X z`29vaHr72q#~je)q0R+7l$Q$l>ST*Kjih2{L<=ZTT&xT?MlCya#P<-Y{}lN&f5DPQ z&$u;IteOV#1;5==5IV-PkOnC65V^us4)osfb6Y_TmXs~-6xt36WiVkuv{#JD#NH_E zD7#-00(DyL%+#9YxJsl{xlU@fQC3*$tSBRcRX{IhKG@5*eT|J^hYI|F%VR%P@gSTDxR0+uGH)HX>4qKY5&d~ zLI2Afl?&mVEDr9{n(0oz^aSEBwhPkXSb-Ollzz2D9ix)yOhs4eO-Zsv<#Z$z_ILVm zAMB37F#{1okHI-bMBgC7t@=N{OI#Z#EstJb@byyqW5hJdzB+|6sd{LO&&HUa5L4`> z4=bJ9VxYvs03NG0oJrmC0B11!E>;0H^&8`vByN`p2h%D`gGw{o+#KFxKUADV;HU~MFG@@s=< z1Xjhf|J7ovO{(CI-ut~ZPio0#w|O6Qx#~KhtJN@&#F; zB-davUmBK*Magb%YVN5KJ1B_?x)*5f{6Ng_Oz%(96xU9x$2pk_P6w;btexW+%y_{h z(Ygia({xB@KqG%`+#o9&3q}@Lucg+~E{=KUMN9ABI2V{t;t*M}7eLksJN{UVLZpaG zLYN{n{EoAzB{HEL?M1Bu#Xzyo3_N_)vptKjT2%(}&U@xQH|lZP=vbJo_351r2bZBa ze3cp*zs*=Jj_T%Y@P~b6I0}B5-ZjOL%j#i)oaK1$M!E96f4;M3G~@b}WeR*K{&lfVbL zZVg|pbw@6mxWWcMi5Lqrx1@t{BFmB6)glF3gXUNY{=daf(%)Fg%I-s^U#2lE-7mnr#=ee zfMS3{iFAO0hJyv(I8e&>H;5p%<=8J&+DUYojq`I&0Q8*}8P-KNkG^%NKbC0Ci3^CB zx<9|?M2Di|f&~iE=ZF$)$&@bgn5$mujF)`H+P>;MVJIs!^=)B<%XHg48>_v{XbRIUGqH2GY7QY9I zkj%(vIbGP>UAa?_)mV`pIXOo7kCG!Jik$nHt-mWWf*#4c12XTn{6SAR(+IEdYAk|73Z%@b2x(mfw3`tLjxkwQv)VV zzTbPRe{h3&VqmmSjKe-cjPvM=S`U|kTBtRIBj?mMs;D@=gBp|R+$S2WGFOHE0AYwA zOV{>1IChY41l~!@@t4V7zm9B?TUizLIAZ%UfZ#;1PJZ(8Ab;G>$KWItcJb|+?ZkNGlH$F$&w2Ur zWHndf9PjBtx1ouu#>U4dL>fm^ul1=_ZJV$ABdM6x=0iUR&@nj$43MDW&$IykDC9kS z^>rXlyURb1=iSir#q#A!t>!3f_M@wBVO{%QD=4K`P~g}8D5fm3*O-j=*aA&74EYD2 z-cX^*M->$IHRMY9;jSQJgN?KxcS3EvAG7Cl6_ci9GxW0ZqtD}_YKo5UH1RZzx=t?; z*g8H?jj8vNKx|14m7PJnpBo@n5C!3320cUzM-mb=P#_y7Q}qiC)?N$n6<8l)lnoaS z0iZ)wlbu1-5Vx#ZK#Y^x8H#r_8@hX35RMO)qP?*S!3vO@^aVv z(SYN+uKh=4Cs3oHvdDx%0t89uQcb;a%I1RXaRgEITG@nMUHqJt_4%Xb& zCODhlFKp(&MxCy6Bas2@(YSy z7x@Kt26UQAi~m=~`VhDR{C@P%Ixaj~2d=mp8g2VG$D!F<6*A}1>>YPo@9~xRgU1;X z5(F~|X4NNd|7^%}oPEA2gG^?Z(|mG_H00@MxA3zNty_43L;{)3T0H4I!lW*LCewQrp$EeN3jlBfC8hmG znIuJl#guAn;R-`|&m!X>hRWcPL?mE+nj#O?bet=Ham7}t4>-o_d10z7QrY`sAw-FD z?aBS0K+^@G@Oc0DxrE*$BZ+e_btScwe`762N!Ws7Ab;a%0wY5}{D_tr z!eg2sKCnz1=dR)i2+l-IU8o*Dp;8%_l7q?uk|Mhe+?)z`AZJTth$l3*6A6Sb1nVL$ zLJE*?NOVcBM>bql-((PmgT&!^9WN?M9N-G>0sKS*?M3AdB}|?=5W-aEjo<^&sFOF~ z=QZn`eRHSIsh+Mg(EMa*`ecChT@3m_jB_!Pl<-`jAU&k4>fh4NV_9}i;(s^vrV=U{(5*?6gc7k+m=Kb3z3(l$iNdT zXS3y^uQQlU4&LSNQQwrJy-K`+XSempNRN%6H>8cJ$KqtZHCFcxQdPc>!D_DEzGE9I zH&4*6cu-C+hr7u19nWXX>CNq5v!8crN#JF)R@q-y!n(@KcrO=jvbnx~!8v;?&34=M z(|n#sX=aZnvqleRi+bJ#K+bjU| zn8vYLBi!F zxHPHc8_rIU7j94!?`BzSjS(R$THefCq47Y@PPL`MM}3#?w_3%+R9@mGps#?7LJAbf z$}JH%IoIq4$T}cG))cNt6|1{xdaw$NJ(>+Ap8uKkO{pa5s=DBH z9QDGjh7MANuw95my*>a8y7Z-VM-dbZQ~f=rZ5X@FhxtHITDSS)ONWbE7l$>@Hg-m8 z^j;E9kRB#u>AbTm&w}et#ovr)sZgULx|Y2m&$BG|E$3!F4*Q`wf_}KYVA=}WFQi@g zMh5?%5spMkCCw)5)h3&DMx!AdJ}(ZZql|CW7XM#Fd?H4tciu=D@rjX9sw$=eScx=lH}xDrS>nlW?p6jDf7H==H+;pVnB}vHAsQ3EfI(>S} zfn&mN_Z75Ei^oBg{N_o=sp8+0wi9BRt3gB;0&?FT=%ueY*xxmc2pdxSlc(cK+x1eb z{yRFKi!`+*wadzxYI@=~)HVebW?w&Y*IyC9m!BoEC%7~%KYF^(tvY5bPZy*-Qc8`$=U?R4eioxm8{SSP==m6UmYJZJATbhQW zd}T&Vh{hBbkHXL_D<24qZYCN?&u{++;92@{u(Aac0i@w9wnsH=AHvdl`I4!CGZiG2 zhs2b|Ua{xBkAT(z3VBKrb|^$CdM?Zo$+_nd1l7MuiQ+RTm7wHlh`zIN-Tqf}>N@uS zXyG&GS`H_lc-pM{Rhhv78^-Y%Pd8lDSj^{MyTsn;wUPCLif@`*chau} zzWkv1A-%76dOttjava7ef4UDl-0VM5#B{@pe@fw!lDYGfvQs$8tfW@~JEjg8btn!n zl(Xa@*2klBF8MQN^AnVd03>{YmLUmv3WH5N-8VVetNH{Ofq$h-0+1M?c#C8-6v&)p zeZymWW9j2kOx;cb|_Ee78El;0W1h?D9_DHp2ooL|ph0uJmxp@CTNqgOdq!o(x2 zvJ&choh;dVLer6M*yWKl%PCu_9HY_iaY4uR#niuh1RiI%@ywm}%VD>)UuA8cRBy1&TsLRZ_BC+wiR3C{CSzR z`R+Q%mA&HDoZSPsq`H^~8>6~+6xH$r4cp9w*BA;?|DOK@)U^o*E(vC_iScBDg4pDN zg;gp;wox+i{b3RRZ?6ciucM(z{=bt!@Gxz zWK5-ddi8OYQ`D3Vzx=UKx&1fsE6-XASsC+W;!shxRg-3RZTj4`$Ntum^oZYyRk# zT+fTS<{IjW&lw{rDU%=ggP6j=^vC&Wbv*WXIODw?Azp8``#Q?7s}uau!4&_s%#)^ugo(7L?>DyT8@|PmN`Mi!U&hnA1iR~>VcXsP>!$t>r?B4@4~BcsjeJZ z`Hle_iZ!r?(UewVf{AO@HB+~N?E5FY-hQednz9S2)v?0gY=H)Hh%|1B2&qRSqrzI;gP~q9sBW9WRPkyXm7R zzft987-CsD+4@7A1e%VEZGeBguJ*N`eORXZdF;~cJ%1C}8*D;XRk7)VV3_mfPoSvL zw?a0e;C8+A`c`0H`~05;O2x8LDRdua<<+HtKCIBB1nxj13 zh$o8BP?3$Dy>iSbq{cNgdr2g}v?ZQ|kJ}sw8W0E*5`T&D9edj)2hUpm99ze~VD~j? zHSw?&=X0@XiRCisIzfmM^7o5G$^`vjD3k7E(K@Q?Ypz~F^Gxyo2Yv>}Sar;xNnC)tQE9X-9gEqk;db(ia@SU5{{&L$cq$sJ7Dl) z-A*a*`*t{En9#F6Wu2TCv^8}{jof9f~MZ?Esen8&(KgOf%@6Yx(u4dT|Y%54#JWtAVs z=UWAw{-1ehKC;A#M#IrRJ0Z}!pGD@2WW-|dybng=-pxAi6k-0~8Wd4Iq6O)FBvf`J z)qxjjB=z`_e~CrvR&NubyO~>x*#kiOkSdqkW9g49q<49Me?mv@!B|DKetn%C=G@xO z?bT&mxjSuzPwM6Ys>-G{wUQY~^pq1W+W};TtbnxYN1v0AW#Xx4U|L2gcg{+ZRM!v? z8*C@gNuH`Eq1p0qsmm?&I?h8SUB~x)Tx6cM_>v=MZ|hj!yqeZhYWe_ZKSESJrv-nq z-pu$;us!3=_k8Ut<*J)*`%w?)%f1J3VwRScK6e@02OwPZads8Wu3pco!7PoG?)+TQl%UW!{`s3l5>v@}W?Ljaju11>fb zZXX0C%KgVO2DlGp=w(e6o{m6iJGZW4q7&T+|Lw^9WIPut-0JOvlqK&+LNbb=>fG=nfe8R?S7Ufp3CnZ40EI2W7;)@*1kQ@pn=YT3jz~=+L zeGL#|qz%^yfOgPUko5$j8iXnY29xj0MLn60+oNw0`2l!wzGh?ZO<8SOPLlIE3*7C4 zq|0#l>_$cpq4&6uyC)pOzKbHSDz(AwXJip&f-=Pa09|ry3 zT0fWEhfrmX9Q(3d%Tj2GbmrtD3X6FGiT*Xa#oJiCB@Go{!(fF{%~^d*ovtSvrEzP; zuPeP@+;Vqnqi{~HTmGhuA5MM-3opS7vMjPBQOB#yNJRLg!JQ_Sd+5DjmAkdpphb%% zAWBSaj!VBC1aRUDvHFkPMBzx_sR&ZD^WAL#qzeG1)v&4iK)=5!KJNoDxhK4VECZCdr8C5iY!<2v=?beFdl zEc`BNiWxE+Kc^2>!k%J%HLBljJLIbIHRvTGUMI|s>l2QX?8-gx1-1X^*w3W_qz3ZA z;E_61uM*X}cJkk%2PJsl?Mg5s*f%}}wqIfCdCoHx_irRMDhy4^usxa5Ry)CBwkEve>%*(k;<=4H_ zN{k%KN>R~~w;q5SNCa_dJC%X0cZq1u?Z)%HBpUq-i!^!!BfgUcwT5()IeUk>7y8h( zUaK>TdGiQm!=%D7L_}zuO72Bo7M^&62q^b6u|_@i9K5Am&;jBga&pBoR?uM(j}>Ep?I3D)^Kj(Z{0c?S zYRTW#PWzuN)5rn9Zv)}ZUhZRs1nHyuk=*Gm`Vd`jZNg{!>3 z_6Fs)dkfdMERD(i?L$LL`){^D?Bn$T-|M>f?Ck7kj6L)Ef8#|;#YJA^NB$Do$Ts2H07^h`r;fX4l4LWRSKij2 zV@&l`bQzGm2eq1=df_1EB76qn0zfJNka)nnL?N{vlAae8T&B2K@8e6 za`XG#i(;0=N&>GYuQ1tM=kURD zxz+ANskDkcMs#a5=W>FoJf>%9-?5xgl`x>>JI9ir#((Hu@#}OFvsQOz7>7YrRp=3!Edyw}ts)Aq$ zrM6G5!&R&_*bL=ANG;j)Rzpg*i03w#ex;T||BMp(^h398->odPmW#=xBg0KwVCxw* z*BOVtgSD8;68;<82L33ZFmMp^;qiLD&ixhmGsWWaZ%rnL?_;Ib=tT|k&cRUXXJQED zQlJ=f+=269W!1@10g0rJSc<6AfF?fH~NykI1)y=CgDjP*z*sJ%$vHwM4bZJf(=fdRYI3$ovmF*1 zV2PK+ps^4nqQ2r(_grvDn6x<<1MBqhN#f9^_-^5rw!IX&j)e}H6s^mvKzhVm-xqhS z2C8~S7%?0Gc1c3O)b@hwiy+8aIDhUwP~1loi4aVG58kC=%X1+|2O) z;1fSD;0y|TH}Q}|6E@^?A~K=eSjdJsvQkLdICj^Kdjo@+W$@07qRMz(Blhn+D*6f` zwxMJ^X-pj==(P-iP|zsk9NT3_ZeZsAxGW-?^?)MJ=4B+aeH@zQF8lWU^;ZS;?>b-ddHPmnOP$lO-}2T^X&2#S(N+I_ zH8Ja_8O8l}kbT;Gt4byGTr@JLHqVm2`(A!;zw!bKAT8NeKBd7z8AyT8b=Z@=R| zoNanPZRTbD%swD%{S`{rc`*0#5-%K=L+J2bmuutO>p9xqrAmJz63<#$fi)%!ja}9 zwVQYJxuJyhR1GAYjRgvcY-|nTbPOGKi0Qcol*>IEYBF4r@3gyE2yW(|sdoKZPPSc< z^w6i)gdB%7T5W5T5Un6#Dg#M4U&6j7hXX1*rA)jT7tLB^gC+M#!(y`GLKKchjnM=I zxWh%y(py2pp|y2;`^vtt9iuHFChVOvU$1+`6KA z1H0-{^jL3g_e^U$r`h!G5qS5J`18x)cvqn48>PHQRd{_|mOyo`p8vjU67r(*I?P@4 zWMi{fe2gT}hjv4naKnw^Mstk#N3nhpCgDr$%A6F4#lT@O?A*`G*=@Ex+@H?J;PYoA zpZq^_Ujs;n6Y_K>o(Kum1X0rSnHB0RFNDc-jD;?TFp22ceoXpSFFvOvPoH5%Ws5w) zYuZI4iDH`;B5&I`aX6@=wat2Y&w0^b!Y-M;8>>^i=X?t1No`(^a4*Utv%>(WpLq%> zyLi@8q_?ZmFFjf}fAaBNrtI==R5>tHv5-Z$$CUNpE@65Pe)HXf?Q;ti; ztf`M~phd~LRVdByH~<65K%XJFjNeQ3f~TR&^6FcBj)SUg4bBIY6uqCp=S5R{UxzO_H2$p_^!i>~82 zU4Gl$PM70gJSykXt>mg_KD~*r`5$UvsGO#n!Z`e65SP(Sh^_1UZKPg9=XoAEdR1wm zuf`1g>ox11;rEY@F9)f?Bz>Pqsq{=4O(+y{AZLGYIFS5*+h|^yUkOo112Vrj5XmSf z=L$(sSMTa4$}FZsP6D&V8!AHO>|@fINo`w{NyKF{KWUqsphh`Zjm;{dZ1zN5h9kb? z21f+g8W8|GvHux!v~%u!x zUsX`$px`wp^rB-V@@;pk0WJ*I(4QySn*{K1MzOcwsb~VHRBeikh+#b*%mKtmRPs)3S`q<*sc;fFlx(GWzs94K$BDeILMD-PQBQJ- zr4jFxq5oQ28TKGq5PP5ixAK(yoDs8?{U5{C5a9m4E$-tuC0oa`=0+AaY1bCL^hT%o zB=*hq=RUaUZZ;1acUpch+>_g#UF2yiUz2`E`p-qBz!aO3u#bECQiMfV6;Ek- zst4dYO<5C5$`)i~W+(b-3+XN_cw{oB={z@KkpSarCxjR6l!gk!EkHWkS8fYC1)q20n z)VPtB1_yCXd2;`GCh}C;eR41djHh9~(&;m@NK=dE5W<2E`|Q|C=QVg}`v$O)%4UT=zjDlX<%qkE_?eHT8oI9z&uGDl1Nm1RxCEF01bLB*O*=NwxQVa73VGQX^~@pz?k!eogD-l zE)X7zSg;M~TESSKW2T34XLi-=JN50H+JJr-64jH=p9uoFuVpM23-5ol@DsYKNU1k_ zjVdub54Y3PwzF;fus4&~i*FctD^x8f1L5Y$LR}}fxF`4$29G$VFYMhflL4U6Tuvu5 z-*mLg`^|Q{>m$Ri>t|7;gW!L|>nI53vKVtUPJ$Y-ci<(!Q{ez5esV!$&NQQ?rsLAe zoV0mlP?Yq7kW&GCqK;5uhoFAOmm{|lE2-c{$pbJt}+ad85N+fX@3~NM9s6@q$WJJ2!Kq5k>N@SWuMt z=C4#t38 zEV_+LR5cOIDNME_8h^OQ)6_NvkC1?rDe$9C3BuR$LslJlP z7L9}`Z0ZNysEQmvd>Pj(bGL74Hy;_l+SzJNRtjn1ia=yzgVEl0gSCl znKT*mY+KUF8o${l5b9NLpEIxNSiscsw!2JBb{^wyAgPXR%iTf!6g+QUr~4ZqbZ%<_ zoqV^pQ<%sLaoTt4(}HZht)J$@=JR@9HOtv-vEO&yazCDz_Kg>xR12@!Cu+mLfXu zQvdY=8bb;37>qh$HrGR>@V&UTN*K4XV^Xwl|G68Rl%KNhAcM9-Yi-$5fn*@r2iTkj z{kqTZ)sS=v=49`h-^*mse8@K>a4F+u@ zkhUn8#yEeC*LEpW2`Yhxlx(z}pjP}{V^5Ue7GqOF8N`{#$yqy3Kn!q?7Lr>DW)tMh zW`&USj1%dQ7`@#_UupY?W$>Rq`#8rX@jZ_gm9U#Fy?%w7>EX3j8OqhspvF|tm3f`3 zG<^sR%6h+UeAJHR_+01Jl3(;@yWCSD&1Ojh)ga9Zeq z?h~iU7P+WY_kSVfe*Wl(`QfCyDHw`2H5j1xRg~C42iY?HRm|AO9%f#+!>?nV1zTB; zyQIsRYlSK?p7-lDV6Z7}e|6BzU5RYoPe{^aZP4%*wk^>yL0qMpM z=4Xo}Bp>nSNf8`eBbejN07`>R4F>nr1w3U;eIUEBztAhHAQoCR#$`W?J30&H4)W39 zM{YF;5$0z6ZU3AZgcR~fJQU{*sW+@r=o8m$08bs9VhsrMYYV7g+N-uQ4LM9;`g!bv zol$O4hRRki0U6%HVSD!GQ3#Ok=oll>BT)BQR4*ok)o5Z~#lcn6tv~e5Fq-WbcLek9 z*IHL}0$t239Ns( z3KoY5j(NNzlx-!0=?NWQk1MxHjb4;Fmv+E8+qCOjdY#my``n7km8KfXhPaiq!=)@- z`1*dP+q&{;ckhDYR(JmUH|i*5#rGIZL4`RE#cuO)a=44!zxe>$%zOWh2H(Qg)a<62 zf_#R*4{7f3W`8&o9)GJsqyERq$zc2U%~3f0r^a)-{;PZ+A;OJHk64N`B$hWwzKVVB zTMvAw&QB7e75S_V#`zcvBVA)#oYYoQ8k@`HZyDyxaQ~{oK;l;UH7zCzwD~kxM9Vc7 z^Figm1Ow&U3J$dzt}p*oDyC5f(;zeh{VO#{mgSGarJC<55moshB^-pZeGIZo2#$n9 zs)C7|Nn|8tX}s?wcCEfUXE{N6Vkjmas$cesye`~O;2Y_X1&gc#Y5!u7d5n@P_nLHJ!l#g*ZFrGB7rYNWd{{! zvv~LCOw3QE%uyI{m_pCmO|?gAa`1e2nU6A>+K=z$1jRDBMI@uoqAkfq0EyIJ(BBu?285~xkJH394i3^V`!9@iAwY7Vt^8* z{ZA93y!yfJDa9giy0{_C7BzVGO1;#FV(>MtbiiRb?#BZR>Jli%WXcRWSS$~JERs3z zVJLYS6B$mqs+K*tCWKRSyBR%o*oq;m1boutf{o=setibqE&vhP0`P=vppI}M{?;xE zc4c8`VZyG$MB-a+n$E0|e5Vp3n+%TjUGbgpM26;?XMy09<1hbXOG}-ljjwk|26uxnO4sdNFuF*QyuuWr#C~@HAb{_&uPt1-_7Fb z(=NH4Ml+kpf6I#xXI)*z)12O~_uP6<=j%_C9v+}vohv0})Qamm+C-Y}YvT@wXywM5 zHt9Pk;h$v4a}PV5Pydi;`eN#zG=vO9r2piVO&uAk;>|jYYpy5+Rz=M^J!wc;PF*0; z{E6CODzt(pxlNtbC!Q2WmCTvbp}#;5-Khh9M5nLL0&r@~m1eyd`Qy8i= zr)zoQB~p(q-D8QZS5;}Af+4DAZ;4r8R$E*9&1vZjOZY7aK<8Z3fB;^@#8m(qkTW+X;uJx# z>Eh&%0n?=k{AGYpLD|BxUNyV0%nesw2h~z0W+)6rO-<{Ac2C8Nkpg4}Bu9)v+dMj# zOyL`_m2`rHCy&)H%`e0D&gT7TNK#aXqg0z-tZ{%!CzME#Fnh&Ht)edepsBnhJ1(a19&_VJl@1P(c4YXi zblq71z*iqmPpg5fGr0n1a}A)oDA-k;7%Ca zF7G+tsk%R4@7mRCJyxrCH#A2io}*F9xA<%lH4UfaGHIR;qBdHDM>uAXCDT}sE|8M*DIH$ zDY!)MhX;+lgD8bk%C=nYg=DL|GQMAE0Tl`YQ~oW`z1-!x(+9AHK;hziwrtM%AM8kz z6#_uW(QjHR**_FyRtmekKf0W{430>J$@1&{WjC|ZSnzEzl$x7EaIX-0haIL=U-;GS zOyu{htE|7e2V0EL3@WE|qD(VuaeCI6X`TU_-wo7VUg9zdCIxVXIQ3($IDh<`DNQ-| zL3!M)X$=lK<-TNhd&xEBvplLVNKP&bMhY?P@a3H6F@iu24I7p{hjAXzMLV?xe!xFf zWLaT8E!F99SWI~xWLfgNpQPsJzm3xsz-~|^l35}Fj_9cP0Bb!UWm9CO+~(Pl$Cy>r zJsg!U@~WGWu-pfQEM9P9<6Y^!`wLj$Arcw>5%oWpo&#btpB%9d2+32AA`zgnM->eJ zN#&*cM=J5THOZuwKG+m9#n3uPLETs{;N>f(Nb+!Fjsn?m{9CDVkHcC*ii(&h*CUW= z+>uU)evYZEoG2k3eG8YAm(H=vDQp6oe?Ep*4W~3$qlz;hrKmOCZ>lcNoQ> z+i3ie{k`172Q@8uwYwP2#5FGMJey+Dek2YOc5aEtRWzybv#dhGiRmJCgB$o;cHkEA zg8We09`#<8SIy18cV)Qh8^u79Huw60JQ3p^;))r4WF71qi_Ile-Ft+DRo7}rHo<2= zVPG~00Qv7&Vp39Iug~qWfiE)y!^zw0b1+O1%&~JhASi%JH4WDaO_xaWyL?$x89N*? zwhL2)5wub!XOTKJfTt%*4+Z}$-7r2zBPxzaL;cUx%XI5c3Of0&NktkTxF_wqEW(&& zZtHDd$buxjj<7@bZi8uaXc3>HE9_!4B9)tiiX=nJen^52JP$Vc^aPBoMD)`Mtgoxe z?RMbq_^@W!hWYDl>*M7*ce;bKZyJLCOLGmk^}ar%33Rz?F+$~HJcw-3Z?sgZQ8jdM z+Q2Nq>cTMOn8m6*Izn}Ff~B*E-^8*PZ_@`gzw$t4PW2cw{fnUjP&22`23l=4Ar{5f zcfcc(#!0UNxhyL`BBMI;=0f*PHRI{SlAI{Lpdl8ZV701V2mcO*C)H7Wn*Mu zG0bx4#-RT4q<$}UG3i$+|)*@)HH8a_u= z!_dA)poS}ZnwhEh`BT4Nf8n2{fxmDy9~q$MFU3#J;GM{aG*|oKYo?Jr^0_D@|5#1V zZVZL6Ef)*aV@4>6`oTDwH@}H|v*IKTOIz$s{6XobY}vD(L4JZ5SC*)VOaUY`)c?n9k9DGISIXBbxv0{aL14s4HaaI460QQAFy8eamLZIF zMu;Xqa8dvLFp|9B_mfvIkzx8%J;(8x<~2MbQ#p*bZFqoWY#)aRgp;%Rt7U{c}(HEDpinUD~ePhz>r!Q#V=6{)zsI}V^B(>?B1P0y4JL_c*}KoP^p*Ab?M~Lye-Y)pt_l?DXL98)Ads| zQ;d&0H{O@>GLvh6wH09k3UCUPjvTbq7*fCE=yy>mN@C{ME>Z#Nj4J zShhF`bn@}18iXdZ4OnTt@?AaJ6kmcOMb{%;q9YF$MmxUD?ias`zlvr^##9iuI9>ik-jS`+qH&{CDG?fX{J!PA{CYXV<*6>v)JtY~?aMDSiqh z+QD!1bh=@RN=lbYb?YB_g~R=10U4jg@9pt0*d81V@lpkdieOeD1s&`R8>@;YBYwsT zwFkR{e_#uPrZD7+4*u%QwHeE6Zt|w<)R0ATTI`33Gq`061cPU%7yf8yBr(27 z8n-7xweM1AcQ9jN(App~&Z;5w*r6t~wys1BgIVSnb?j>7@E|gsxXjG|LRd>c(t#Ih z@C=B-4923gqTs8flXN~g9J*aelgViV0_e{D3HF0xP2M$G;G@})O(<({gaDeJJvq81 z-eE(KvJV^~*D%wW;sMJ$aHy4u-2I$;;+iVqxRU>|RLw!Nyff5F>f}T;N)5t>OB(aY zY#rm1FD8wz>FO#rOeHt=Vt;gwAi(TgYsTURGmRc&$~I1w(PRnswvsnSwBS0t#7o`Y zXya+Mjuif{$M2y`|Du-{EWwLdX6S#a4%k1ZRam=^j}{60){I~;ogQdvt@uaHv3*z4 zhtMsX_XqT=nj&nW<5tgqm+j#jcoltgxwW-*Fpwk|vhsh+y#zHL9K`S7O`QcYpgkK0 zx>F{tV;|;@O{~HKDZxQM*&9My!YYROCevll=BzbXONA9AaJXJIcCzIy6Vr6(YJk2p)Z}!= ziRnsJs{D4#Z7(*A@mke5rhtvu;xJ|7e9>5+wxQCgrCV)LY|_;D6n^Dc0WU2YO-i|DQp8i{#Wek8E^3v5u0c{Fo`H`t$)7bb8G{ff;>2rm&-m&s!W|v<-lM@<5Nr2KV;BM6t5fSlY zW$QHSbj{56UZ3xGQGz9pM14^HCmKK!g}??!6vm5jt2k|&4Gv$poM(|Q-Q?LDqv>h% zf@u6Iqle|hE&gO7ntn3ngIWYX&{v5F6)g{7t4@AWVN0gT=48)*4TLD+uhuRd)I!_?)^a*)U?yT zzjPG#s*~z#`|UcY>IOU;3}t&9HQNE1f?8czE4nxh*BnUZP@v{L7M^1+vZ zkCTQUVJsvhBtPKk7=%rio|Xos3jgv!+~WU3+(NLT&&)TOmF3XHhRO(R!~7^YKYz`} zi^mt431dydeRa`KMV=g893~E(Ze1FvG~ScVZAMaSBUmuj^Sbb`m>n@)wPZ({!O!cq zpO6HLXNzLhA(S`T*k(soGNhKq7En-fW9lO9r_(rvpV*~duYlP?7PH(^wEqL^v4WPQf8lZ{v5#j36St9-M8468{|5D4QDO8mtx+k(=yPJ5;!iNw1Q|1!T*p6}F;wk!|NUw=FC|U&rAxUD*8G7G76rEOAw+ zcKPL6eoh{H*tnLr4rJf@e5~XmcG(}s);3(Wa zkLf*$?Y#XJN5j$y2B+Wh0Aw(Fod$n;D1_#zcVm~$cgbPZ|4Jg5|5toj&cP3zjr0kw z?_uwy__^B|E(~0U^0S=+ABMaH5cKw+bY548yjp*x;oly91MuXqnuRdO zY^4PxLBBen75y)J@5uzeAAbab`-{zP?;DH9$4R~sd`t^=WO~D~FFipqD1FN9(JU6@ zbZJg}K(j2p?5(9<(bk1!Xp1sQt0%d1?`s+=Q&Rcz=xTkl;}dn21sFyr3fj~v*$6zh zlc`oJwA`|jpFnp}4Q;cRE6@#Iq&;80*2O19?B=pz#1(EAM%mj1k%x+d)H1WQa@1Lidu(VMspT*J~cH4pDw*`8!s-h9#xc+ng!YUrQlZjWwVH) zL{hG~R_Je+U<5*XQ^&@(;$*!fT?w1N*=rZeqN;lSO>avG0UoT(>~Y1qqCZVYG8P+i zIgWJZ#ZXE*b8zgwRY=p35VHC_Ex^^xE9PmCo&ZcjS4d)`wgkPONaVCgEwfEqn$~pH zoTR&$^x*#Ulf9WE&3=<8dJ8CQcV4jOCgbfjnER8KRFL}q-MDga3uL|6; zfJ85k(AW)yNOG~Lm(7p%6U4&aTm;6>6(S}N(m4p#0Q_9-b{}5pcfY~=dBpjW6iqHe zZ~0dGVCa1k2M^f+^!2_j3xXasj8J54+%dP&S*|&8d;!Bfu;+OTew#p)8p-eFdi0_9 zBi)^(iTd;)a`(F)&d?S3{j)dL*A0-Bm#@{RS(^OX-{Z1BvEJ?|;)V}DhGE|T>^^8v zqwv2+w1A&^e7X7k%D)ev7oZCuqCP1}5N7!DWu@2G^V_#UhdiwOSR8Wle*fpY%DVNT z_RqEJ$ui#O>6^6D$35ov{4WYxP>uvGBnKksc+?Cw0Ua{-@(v7+0%!$U>KrT;3wp@?ZxreJ)soj`Za;S*8U3z z)%xHE1Jg@fSfP*%``s`9UD6H|3A+<4S4t1{;)Y`R#cuf_e&V;M5D-u}LTbgvr;gb9Z^Vx^c{?ISlsbjX zw1VoX;SJrs06niMvRnDgH|c!m1f6Q8J|}e79Lbtu1HWeDNfwx7)tk;|NdFCs%$z?G z1cd+}uhVw5tL{Oa`>=0bw@Qlecn*&{k4mk#ybnvrTCx0xKcuqFHAQkGIxb#+RGj2* zltAPDCfV$SHmQD1_Q{#Wl;23ixrv$TI>Wd|t?h9Gm3o)P^=b!cuEvr(sjE(enRc>+ zE@q?acWt+gK29U6RxzFXR6fR`tlgD<l;8nm_ zKMk3m{NKEd_A0*s(?Z-%F@R8%GhQ(3ml_5Tf^%N26C zg$b=+pKg_{CAA-t7n5Q%rL%e_^qH2HPv$#YwbLBMe(GVhNg4`xJ}#MTAgxmwYK)&S z=L%^=VeNlbU)05Ib=p4&hoq&lElXHe&ui6%Vh-rcBJljk7IfUttcu|op~*6;O7IlQ zRS^vMHOF(p9;Dg0BxO$Opn+1p&!{t8j6A%Ibz9t|=NrEV2>-3!gU3pw;`g`hHJQg= zuGp^r@548Lo9q?;YFQ@WWrZzOA(p~#zdtBECSz16U3Oyp{@1HsKE&4{6ebXL)gcCd#NoU^mm`>BG}-PYF-pM!Q~~YdDcRAi08jDl-!d}Cw;WJ`Nuee9t zpny$0pkc$N%W+p79}WITbA>jb7b42^zP~*e6%`E(3|wDd``%wD%pHqT_*;Hwiv4|b zlqng5M=t1f)qB5nbD8g$&){ARB%!3F{7Vw_ zw1&ERE(?E3y6L;wd$gbYd)dtU`pGw-Tt-dcB8|X)c4nrlFzqrLHhp{GbC>69s&jP* z6%ITwZ{mZN96=(S{>Ce7%-u1!;*vB^8e7ur(WBD{jezXGE#qWC`vBXB*`JnC5Kv?@ z-Ha|8#E+_<>J{SM+$(WT?9AdN2eRXG<}P0Gr?tpI%2;7&N*Ea*WNyCt<}={k%k%4s zZNCRZ18mmF>hxB}!fcB!AuvSdnKW!Wiu%_gB7L*mE{|KB(G_oA%2$o6f7FmrlF20( z;o!b^2SXitSRm++)@zN2hWH8&SdF>V1#G$=1-JP5lujM?MaAFGntV(9+ucO-#_Gf2 zvy2JBye(hDAAT)cH^(x-!}VkL4y*A&(?=B5Z~iU~*wEkBhyjs8?37NiYo`R6Gc8Oe(BiOQ)9xw` zZb^mrM|g*j zXvnZr^vMD1Tv^hy(3AB@F<+1n02Ueh3?Mr4UakkWJ8VAfgrc)5y<#91Y&*VP4c%`A zzHdI?e>Ld!emc8VvrKHbt}uITRj+S2Ow$R`f1A>PJ<&8+MSZ1j< z|7Py37SX9MzR!?bHBW=jR@IP0z$o%0kUl zZqrHYway>0QdTfo$lAk7Vgt$upMrJnQTJ0q5eWp2%dAUoD2=NqsZY z0^RaPL$Zo_Q~um@y`GR{}n_+Kk} zjN`Jx2E{7~rW<-tR}8lRiDYz(B&${I{Azl%gFxd5bFf{E%v|l2jkzNf62GZr>ZOYr zm#ZaQI-`3faeF`Epw&|>ze>1f%q=D^w%KQN4;p$Md?{$pSGuScyqn^3!}f6vI4aMQ zTPi!qrLXOfR3rV>%_@UT>0R zX15=cUP@MHVCP;lPjk#y7}4LlSbf|K1GS|*0+bsN^(kYF~)nyH74pQkNJP8OH+LksrX~eHOjH-Ip4)t8()nh22H0$mv2QD^R z7cV5seynOqq!PVgtf_>ZHI>2%-U+@37GiC#TS<*$$420eich$$6~ZJzn(LM}seQCgdC`!QbnRAE@Z+TvdTgy;CF;8`CUiwJdU3lRo*{sVoBU4?|) z9Jv*%qfy9hUL-CT?`tSL9E(Ocayk%0P6QNaf1}p%dhX6Ey`#MBy~n&qeP#p~F=}pf zIK!yyaO4T2mtL|C^h}Hu1Uy+wN=mx9aRMOZ?+?=NeUBoa-53d2fL)ON3`2*NAG`m} zTh__o3O^*Ta44W~o?JR4&-iL65*;1@LFerR_z4v<|WVU)!SZ*l0*M< zMbjihqH=EGeb{QGml^pFKHSL^1zg>=+1Bd0R4bxYG0H50YQ2**_yb**qqW`|oh*6S zQQB5zb-Q;UQ!>ns&IoaXA`X`2&B=EmDn_>c z4H6*_AkI}z9@KFA@v;hSP*m?mJ&WM-Syi&Jr~(-Oz0RaIXK}OuLqNR02-fRcyVOUz zVzci11onh8`mPi#{*mHoY>R6at(FuUL~Y?-1X34Z;Eg&e^TtyS!W`C9qykq{Ad%L# zoNR`1TD*5fWe#lJ^%oBb)sk{KCQ9wSWUhohcy+QXO#h5(SQ_^abQ{y0ca6WX6iFvq zP<Pf`nwPz*W5sl3P3wOv6b z6yT2;`ZAf@@5}zavD3(5=v5N%OcpvHU-0noH=+L+>7yAY(ik1KdJrjY!z6;qJ*lXv zr+^(!l(e)5da&R$qPSqbu7BVnac12Xul+PF%{pE8(cKJaD6DG#R0%1)Xf8bftrM&L%j zG>!1J*5vTNq(zEBo|k`^jFQM$ME8SN%dj}fsIH_}R^FhR8*ve)BnB$MOX1XTjBbxf z2W+BSGON~b(=JP7co6pf_hRhlDoH6Jq!%R7--vM=&R40zw7(19jCXa|k~E2@NsCHD z7jBnatU0l}lP6j$gKO(d;n!{B)AUU+jZrQpifwhb=JCTYKPr(1o)2GW)zxumi>Z3v z^D^yXO@(N*XM8phP7l13nz|?ffUPTL2LElSuA8|@SnQ6B80#6-&9(p&C{HNA{Ucve z&J69)E+8Pq4g5j?@Np!95;*v!=~8{ARfmNg1+6O%vngS3yr16D@RqLkFHKG&aW+^s zw@q7|4XQhAU$h~leM7<$I6|}PV$9%SFNRpo5kma7X{ef*G&tg*m7Q6t!HcvJZ$D9>v>(Y!DGa<>4z}n zw4FDsmi|3;Ur-CWA`e-X(4T0Qtzuci=Gt+WJamYx1qC!a)U=@Sl9gGuI1F|_m^>}4jCv#DXa!YI#nMKEUz{A>V-DsrJ5 z1ZgR0DtY`ybIE?z{pf)&axO;UhbIJw)K1%7w~iM(D4&$5@fEOx4`tOQDYLd?UC;9C znxn}&8`yqRd-9#QGm92^IF;5Ff5pJ)r`YBJRfH87tEIG{EsOZ2C$CiUI5 zqL-z9QwAzyieX5>Rd-Lb#p4PF+;X$Xx-6RvK1-e~BL1`Guu^Z^&*le?4tC4#_yi6KB4uwY>Pr7!$eEb4!+!5Zjwwf6MTOZh{n=s3XG&s9Mm20Jpf%qD z7khRCX=vD@{#&jjja*fZousWpfOc-F`aY9O%~mm&ojySH5K6QmHWwEw9|-^qQ2Y9> z*Un}dyZ{cx?_KrPzT#+Ds9r8UO@A1AeA*8Q(5c3b4EdL{eoyYW?msr^$dlj4iqf|$ z&4eG&D^-k(U8nJ9_pFc-Ti}+g&GfJy9{uf$QtWF3J)p2ZLqduc(EEY(O9avP>jdNXr*(dO5JgHunn>9f6U3nfIQS=n>U8(l@3%}|@ z#!n{jDDjWi2GB216*tBAFYb*^_l{~A==3|g;$1p4?wyr;OLH`^u7^yUJ37IAPWYH^ z>st$wH6?(+=IF14=B2nK%I@-96f4+i=5TA>;X~|{F^v&YY`sq6=fyrF??TFkLa<= z+=nH$Z&@p_b`Q|2fTUv%u?nEN*xHb2J#Wh?BsW=)Dq!#dlTUi(d;G%gDYgZ(Z>!XN z`$@YbGo-fsKBr2y8-FCbPc3AvIQBS0rF1O z7{QPV*49}k*Gw-|D!iSB6Fci(;$q-puCTde(H=8Lt6h$<0eO;TJMcX;q5q8!Z20M;K&c<) zqyH$MT&K(aL?)Z*f3u2#PFmI16lO?# zf{Y};x{d7ZmI|7J&VNx_jMbJ>MOqRUb7tK)cq|P#U0va))lx{?7r+!wi-$e+nt#hPNrMhnJO+2-a{u+sWg``&{gg#l$qsiQHL=*%o9a<6skpWd=GjepM44B$ z3|^F-97))O`*=Gtqr<8Yc+~=6TY3*2ZyM8i1GhD{o-` zPuZLeR@k6g*|Mhn+O=-oL0vc6MjH?ASWbZDk)_z$5sO@R=*&sr^=7jcSqsFO{rG;siHWjC4gP;L17ZuB8nLIJ8fp!fDao8=acz#UpYpad* zK{IJn*(iy?gqi`t7Pgb~YWOTV?6M~5-MFQuo}|kgMH-IblU57Il=n}Zh2#HKa95p* z!UhaW(bnnLJke8h5z5YHJrY1;4O|V<{5`P?jwz4Xuj#d&sLMaVo4=toJ&i`XwktkvmMoAJv}~*#+{`p8HYxOMc+kj0ncL2+kj!i5v-ul-!FQu{B4IbT^$-IG64B{RinBch%~9@h~-$2ThYB&s=fhSrE(e zaS6Vkp}IN0YHkv@oHSav4pwn2$8rcf)!5NC{z3=koM`H>clh?$AD{I$`vU*FRT@e;hmDTl6{3~FYoYT0haZff@*k99A>vVLW6ram2+%6AuH*q6 zDv^K7HkxR-MUnJC8md5(9BHcK@~CsmD=~Hc`kXf2 zZ}i#OKV$`UEPo{9THRn2!4#$mFQ^{Y?4MmZU#h|gnZ3ymSL^>F7~qQcU0#{~ZSo7+ z!?G@Ou%$9a+|&x2JkXKZvApeh1?QJBN@S$Qo}Rjv zzYqUp$&WrGdPR|d#X@S#cps_-jUn|G{CM7Vzt``cI-JgoH0ZOYvZ-vjd3`rgcQ-9^ z_sGAGBN4peh*&cKEf$rzarJB)gWQ+UFaA$y*#~=ZNWEca+f?h%#X*0Z#0mtwzto5~ z>~aygXEwX}TkOJ}nA@^d6fc+~^U|Ocig<2GM3H;HB)T;+j983YXZeO|R)s_YS`qt%$H~DaZmhYu zMKeFk4BbRS#O`B3)WSI?_?@Qe3+Gx}DVFU>UJWjos(uCEVC54La_0A_Q3MCQC|pJg z!?JDqoJa)hTwJSvc=c8_806=C5k+43lp=MY)xjbtNSjcV83}d#y4*9HK37yx+SvIm z1Sf0kMmo6jlwe!nh-4@k;g3-h^=2U0`71JwWx_#m1$EN&AX34s<<_bg-IA%#oyJgf zLMM&3^-Of?mcbQu&Opy@F5v$<9(jtVFF z+*Z5!wo?7=#mEK=HjI|FlW*5V2;Cbw95-$4&D}e zj>4!%ZzgW>Fikb;i+$%^)UvXbmdm=-To!G+8bJP~i*8~QDYB}Qw@7*yg?MUOmKDVv zVYB@Uu1KRj3mfdMS!s?SPb`w+q46OTCEV3y7vZ_9rYsD+D2oY`6_sK7Xlxm+?3YiX#Z&-eeKOL^q{_W;ufTPG^2!_WhPWEAJ??ZsDPPYBs+S} zXwXMSM<*BZ;bhg3lpN`B{#&ipKt@F5bGpR*vh)1D@PE#Q>%+Mqe=8sp_Uka8O#dx- zGgT}ZQ*YRh)`=wg+y@7Uiw+9)yx8b;{`=_#9zUoK{kEkVgg5ws{^eR7*CR0HgR#bs znB?MODypAT7`ee3ECxMM^nb+55{*UP9!y)1S@LpoQQ885^Bb^eR`})QJKiS?^g-Bx z(G7Y3>umd&0l=yXpUU_FnD>CA!0<0&H?v2Bu~&h(D#tp(2N=cTg<%*PN2n?!%6U5{ zqDsUTJ*x|geo-sE8eEFlIz9uBS4YETr)^}@5GY0~NXg=SYDlnzMs%KL{aLUg*%%+#X2 z9HF=snquc{y4wwXrthhf51Y}&3=G{Orrik}m(ONNQG2z|Zc{(kQTn_BVL1K)^Y)4< z_~GN_iUqwziN!zWaI+H5>AzW*37!HTv^$+g%R4>;)T)l0SS{6w%{h&5BN$>lt}a)U{q6+U%Dt_{ zole?*zJe2o(@wemWKKz&&EOg5)&^7HY@EoMBd49D)4}+ioLt-tU8w<~BqCk1z~?J= zf&Omh^1o#p45)Xwy+&pY%Fslv9Rerj*AT43`k{@KR^#ukYcR}=;0M`#vvrf(r(^-= z=Q&+ZVu)SqnwU5pP1ymi<_0@;xVt9)0?Ft>Z8a(W1Kby|BUc61oW!TRY5nD1Ns{PW z0Z*g}N1TiYAIQ;ljp3`!OfH{vJc;#u`Lvpn5H1!L767cIPDL$320*3Hlr>hoh0H6# z!@=zz9#XdjFan5XT%YgGdwuUeo^$%@`dd%-QGm1P*!(I&r1XU%+e^o1m+D%}iS+Av*fMidjOBFkv{uFn#5>(K^)(@|8v%3sg1BQ)OlT%4?>Ami7m#o>9%=!5phl>s zP#1)>*IF-G*DTz^J<4~I$a`#RtlB<%-nnq+k4-%oVk9z_%yo`ww1G{Q62Hp zh}%646URAV_4Y%3SxKF$xo3o0IDM9r`l!q0^V_LPc0wcx6tN|sY6#plaoK~-=Wfg0 z(T7SFJcYwOFQHwaL{HUtae`=xbuZ`mY98jQqw;R`JkW{QL0d}~9a+{)e3a+G>9|(f zIe0}N^kb>vFQb06+}8x<7>wKw@N0@9{#)8gYA4g^7VQx=o%&A8+J^s5rNwt@FWKYt zw0zOt#dfjf_bg)Q$oG#U-*JN9 z6Ou{YOrW$9i>vsL#ulrP7$I-j zz-u>rWkp+ev+}8i7y#au`=3SzR%w|XGoh7sbnx;Hy<>`M*~lY0LXi}3I-5-5F+bg& z@;Lwb?;znehfSg}E_U`gxM7W&z$$-q6|x6n$OYTtH{~iQNT2_66hN5(jBRK+ot>7N z3IXHV+I)@uU|uN{DBim`VxB%+0O|u73CTvGSZHP@!H!p8@&AuRQMFcjeR+CLw%=d3 ziMj2fwnO02$uO{8x)X$-zVPu8!t7*O=2z*pvmAs|crA1kt3M_RfoXF9^sH;?atoLp zg{H}f7BfC|aoXI~-;MB~jlS{wM;Rv|3Y)yl!3sL8BmuMwx$)^` zJsq%r=&vbl{Jbc_!qp*D+fM3PA57a*kt4sZhM3!?i}+W%2!-5k?@_VtdC>j7-fc&j zgXyAp7<6GXSI|;JI@4k%*RrSiMN?d0<>JeJJMmW?>^zl1_z)D%N4I8Evt%lK7lpcm z2E~$%wwnehP^g?y>{rA7I>lxd#{~yrx0NV*wQ>ND$n#nAL+@+D`%}ZoVhyz}{GF(X z%KtbHf+S8JwcHqnKbw814;D_VP%E~H{PWa^T zsTgj3aPS)ia3?rl$anA*uu;>Ml8O*?F;|8RWkoOy)UR;^K>qDdLKR-)MnKc#YQCvZ zn#0m)u?KxhHEfa~75s+_SgSh?OwbqsuFwGr-)9tbHNm+Tikhj9Q_O0kDBwQEOW z>HOt$%&!>6gGKV*QHg+-JyipT!{`GI`@vktB<03Vb}h}>va}qJmSgTp+jq9H-cLCP z#6N2A`<3!KKkcb#I}@)EWv29iq3~_rT*+&Brpt!%Fri0&JLRa^$l>o3F<(ksEDv?%Eu9=b zuWVz%T@#js=#nX9a$KSt-Fz4S?+DN7`xB26<*$CooLv9I$k(^??YDDFk!PL1zv7hg z_~FdX-W=qR{$Cg-4w8aC%4IbkbPTx3i6<3+z72p}OTTACBG5Zw2a(mNDJr7%K}+uK z@1qj3!(xi?^AmlOla*D5hJ5$JFVJc*5b}S5q%mpi1;ES8+TixB(QWt4BWp&x_cosFnUe5eifWq#%P9D-{n9 z?kVJjoaAI^bk6}qRw-}q#}5Ef;HWw=NA;_QVjLCZ=EipKpcJuE)M0(N9C5XeRQR{= zGhnk^ftc48KMxDrb6)CD_Pj9j!Nl$MblLH0XF$jY64db_AflW-*diM`IywYH1jJ6V z@&;_hmBmDKTJ^(p7luFl{`A@&A%G#qW%2Q`nK!&`)hQ8y&4*5v== z^S$}tX=h)x_6tR*#9}a<6>cD_|ILgU94$+w$P+fYCDb2)iNi));Kjc zDqCJ*NUHI4OyHsUU{h|w#q#DV0=nF{cacQ(&`4xa$v^Fv=CgiXqN~cnuR3xqJRi)& z2HjSOk>R0(B1lI3|1~szy3geP$egv$KbUm12%uSZ!f^r zXi1*s_EQz#gLf`qe#K{Xeeqc~c5kWMit9SGnt@2#hD0cyELfVgLzzk|0ZY+Xu7Kfd zKj$BGD|GPHBQb;Cf7)y}ssT1eeCRQScpV?5)TQ4;nlw`sf4>ze#>~9>qGlmP^d?2Z z!6ZWcVMOu2CaVU-O{2#harm+C1iUBB??{SnNxEf!%?iN;z1kF_c$PuvFS9)CtyoHJ z_b9&wU>6ZItLMm^y9x!uW26Awo(+c3NJN0edUV9)8q!z77MvtP9XbVrnB9k&>~Q^^ z%3bL0yF=$bXyH}6tVd`^i5hO02xJzKXg}7U>UA5M3}(O zDVh$piEyEd!%T3Vf;}f*G-eT{0r5qjGMRKq!;`|dRDezwin5yY>fby)Pet%bjIK{A zM~w9n@5}e9(dIbdW7XtQMB9u{4ct4LST>T}pbf$znDhw|MTU84@Zq+?*`P)(Sq!?} z>`^}PGQp@B{bBRksEz7?;q)xNNDQEq>oQgI!NUtW_%Mb0JjWm=bpBcLk*jD& zWYGjdj!(R4Xk!9_Z!iCQWep)Ap^m;j2@w&${Tipt<55Y1GIBvswd=vucX|09-+Pvp zwQf(xNHo%qn8W4waOH8nwpgpvEEPx8QNEe)B(pjQ^$xoQ&<~ctfsTM~A08b=CtwBO zip}@>dPczv(gHS1zo#hX@!wu<_e;j$wFJNmDwqaukp1vl`Cd6 zISMlJ(AS5v)#eZLii?9I0U&Ck@Ht;=&GZ@c-`yv06!N`;W$X}sA$FAox7*D(x@Wva zjVi3)Uu<|TR4WLuBh66jl8=A+lxVYQk3oSj;fq3)@Dp98_gVQvW#vZYfwfMcUw+c; zau!Uw0!~vXx<6rN^2xiZmUln^J?3td%H`tqA0SsKo?t->W<YT1ve z7W~82PfHdOcu_*;3utja6Dsx|4M>b}ngQsZvvHlu` zX0+gUE834y$rg+1muy%hTY_#!nmIIJv%XJRx zhRhjyvi5)pe2ya|>;V*-Jq&j$tlfT^k=lV@f+{5zGFLgb$O~>nv&4ZB;&H%-C8*O% zO@r9`s_MBDvKlS8O+|$q)>ke!t=}{(UxIBPz}K#YzX}^U1nvpL2a{6pjh@sCcNa3P zV~E?15g{lIKf_@0MWe$Gsu~wO*b3pZ$8e%lTDju&Bt^L$cZ~qR2x~IulJTMXJ(Q!+ zr2&=TmpA2R?`|^jV7SF?xaRb)+L%dYJP<%yzqeuq`5=EcSyUTT391)NE1mpoyecarJGT}xY;~pqO$mJB)9CiP~?iwKk(uq zvj-w+6>K>Q9nFNy)TdIBm?SM_zvO!-DN0t!gr$%OSvVMTvj?6Q{}`Ib&J z*mtW@dY9gP^wKmmmZ#GlE2}X5`43|ZJkViOmFP8vh{cS+vWyo!F)ey67P6Gj;|u)W z^E>S91PnbHPnH3h1;GxX&fGV`fA^t#SteFCck)OHh`$Z5|X z5!J5^5sNS6elX-d`-n~;%!IXQV1J$=7;V|Jam8VY;#>d66Xyz_AxvhQq!UTy_2}+Y! zQyWWRkbUX48n`up1MDJ}qk!x3O{#$zC<=97T}g>t$d#Js;9_q&wEsI1j{$HIaMID_ zFwGK(LPT#`kNypC+1lLv_U#)~1=M!3(iTk`dYJQe)o-t#3;_w$R8&|pa8JR$AgZ*^JTwpOlwMYATp~b5=%OH1FV}UNh$2UeiD4$h9LSa(Np}5 z|A&&V)`|&4Hv&*A->T726) z#&1jDSrq8SoY(;+ukL&$qAHUTG*!lzTJP-FHa=$i>|&kgzUYWsK(8eOa9B_UmMA1_ z8B*4=#@pZupGrw~w@pt4A1*}ZDN0?3ZWabTCjdpOO=KRNZix_i}^bUK2wU#_6K#jK!(1YhP9SGD2BP2FX$ zRW6m^LBFDDel9j+ag|&Q(cWADtBaW_F?&KQH&qK+xRD1|8h6OsV|bqVF#FR@%Z%~J z3iVmaS$HR82+*BFZqDcWh{wFbqSV5A2U&_OP{YbbD^pGpwDME_d4Vx+C`dpt1c$Pa zMJ9Xm!xing^iSE4l<)2K{V-ns%kVEvyu*f#T%*dIJXitOF&&QpRM>}9Q?g4QO9})Po-pj*6k2bSAS4FizDhm)w ziz7sz-Jlv_Mra?=x>FIFV&36TLh>i}{ZyTuCgjsnlN(s!dU|hnvhbwqNm3Z9iQpIL zA;=VQX8Hf+mJv)uZj_ak@!L&Acx{n^B(MJr(ph_Az!MnajNQpVaB?GHE<@33W^yBm zoQj-5J^LjN7FIalDNjOmyn)*@Ci)M|d937!DorlRvcW1dy2#l^u=RY*#P?oTy zaR%-eJ`{5sft$s}Y=Rntq1ep-dTtB0fTX2Bw?2wARCQvt+W^S%%g;7?OM|YKP_s{r zivzR7rtGz zyt3Wi1!nQC-1=@YWt5)=sWIH?*40Ql7XTNuWA2S7&~6wYupRA)LJ2H7OV8md+y!y3 zj&ii4zg|{#8I+P9EPbyfKomqQ;5}73jO7l}hh#bx`dZUAyN(`bL@-vh;`UvDizi;R zHU~a5&aI##y_OBRP;Gj&=1KY%&)o~=rO2rB1`kYeMzsO%Xe zhDv(5?6p>r=VrXm9g`9Y-qyZYk!}#{cg`L7<%@8@)6dFYGE@zFC^uMi-NaZrTCafz zy~-xJ-W<~EWo&1hdCW^5NZh5@N9cOu63X}T3=SL$AuKSR9_B1*j*MD2%MW8^u@kQ# z1D-%=hU{FQz{$MOy9rFCg|AP@fB=2j!sr)JDFuf(`7jZO_$<2K4~Lpj5%?`bkJza; z{mqVfjd|TGv|yDu1tB{hHlLJ<0Rp7(@JFX_JC?KoTh&ucuk z1}^t!o#tyd=D%7yfCu{H$w091urMHu4h6PHoPV6=I>|GU)a0!9N4EPJCm&L_mY zcE*24yFBT>_Vj)0DM~-{&Br(Zn;H*}P5s&2H)!y2O9D;Zex{KOiUQ&1LEZ>Ix=X^N z_F8di-*OsMSuv z?5pk?1OUt4EPDx{5m0*g_YWLlqU>u$LnJ-VkYJ>8PHnTPvXnF{%#pGO_Ldr(P82@79(GUokyGZ7{G&LP`!iQNX-lSZ5a!_5<@a5 z&pN{=3x!ureq`iQf=$j_jSrCEjGfOjRCCiB6hNXYb2hM%-$CQ=00iBdXd6kU&~*;5 z(CX|ia4b_oID(|@qkKiOr{UYV86 zzLBk#J{eVQBPw!0hj08_%Eo9sscR!snCxY-Yp_L9ySzxPSAyws=N9w6Gv>6IRlCULQLUlil@1Rb;m1g!R(1t=%(A%6Y+uykJuL^e+N)sun8MaB38G zd=Dj%$jQRdOY48GV>~FR^>d&D{xiQK$%7yjD7Rl+GRRy*ICjPc=7WHRU9t}CfFQq{ zm#+gj!Moo9_;XB^3q#WAsd%U~FU%h|7hWHfN0}&Q@hL|#U<+c@i3C@jvn@?GSOXc> zB?XaHCLo^=dl^`8dGFf1U!?KgaMhn=##R`TCmGu;=Ei1(B)=dZb*Flb;bQXDQtXXk z#&tpMn=RarZ=#VvTS6c`bVxQsH*02s^0_fzmOq1dul=c1k@**-7(C>f)gT>D6v^1( zBmSD1Eb+I@;%T`w-hs;PVGn&`hRn$Yk_Bxn2gd_=GBw|cLMNitGXgl;KG81dQqMn_ z)NJ+MHOqxa1YADe-RiA&b0L(bHyUryjU!AzBKZ=2(TfC6y`8MCiSjq!Et$%jXVLoK z=WPU3qGU1rQc_*UO|F^}m&COe&dwa&$2Gy1>=D}m*T+E(dJoz~U|2LZ1rqvFi6&k5 z_-us%I2NgHD%c(g-Vt#9=m@M7sggn8i45Q_MnyzGge<^tFa*pC-V0krfj37T6o2^w zhya~Jjx+Xim4?t6_h7%6l%%gv)opTaWQkiHnbN@1h@vLzl$H@W| z;d+N@)HrFEz=GQ_6s&5~HiaXU%98%eZZ_&<6+|>9-f05@sCdn(X}4kFv(V6KIuTFg zLSh|RVRn{uB-qJnM|M_LoCxzepUOZ21;6XE@R;ES9PyW{In8(9!_m6ohzz2WbTrA! zCn6{zFxYR{U2L~ri)sEceLIdG6VuBwfr13~M$c`t`uW{^9q2a1ft4Jo4(pZ5OI^S8 z^;aV6hy+8ol@x}*h~&qNS3L8bih9-esJ?U+t{FZjaf_XYojWA^d~8q3adDR|9{K|+ zGoWFNUH_F!^p&M+M`MY`Mn)?-LC|1u9a`42OdHHJVoLST+Y(@6;+1IhP!X9q*Q%9D zuS@?b(A{>l!a7kR(rWo^{j@R8B2ow-T;6r!7c?nn8}`|(fy4cQp*Fp}B zY(M3PMOaT4#=Jl1@}{f~7@T!{DeHRp8nKevW3#gWS6GY6HNi3c{UceVAhHOXso_#M4c^ ztpv#8EM1ti*$UCWM=Av1=lZj53lg4#(h{;-aBK+{DQE=t2$aP4}3`***lbDDUh zYo&H&x@IP@&df=T=UW`thYxaUpW`nVh}~s`IbAmJDQ`LSH7U%D+x@ZL@bU0Th#6O| z{dr%L@BXiQjQT@!`0P8n`;Ct=p|S6ASMb5=uBqapw@$VllmPU}@?4`s@lYxajhT0+ z&;3BiDh>&*_)SqQOX}@kg-?a?8()F%8Gj-8?N0Fd%!3(*gnbuop!3Kd8z)&KqA-iZ z9SEAp_vvI^h*0RK{&Z!G>hp&ubrr=KzE z(%z4wC1*FVamCVtW+asA4DGu8#nRiufQvAa_jPHNsBwmYD%puLmV6)GS#``-dQ{FRU#dfM zP#i3a%zi&?4AuwXqHm97@~A*-6k~&i-OsTewA77^D1nr< zG;k{_sGF!~{Lj`P=;mXR&F=xOgzG9jp5)(QuuEHlYmEfm`Yj#0slrY=?1%9LA z3fMwTvIU*eQc_OgB)k6}94Wz-eJ!x0(9l8i06OaDq0v$BfP3=S3L9Xh#$0OASlh%z z+;OJV_h>P)`t|$IpA(FMO`u_6VWIm^hJ&2EJV@t+iFf?*kZ~EJkVIH0UYY_Gna&wnmmbpLB4n2S3fNaH=z0ZpICm6-S5D_gzrkg z0k|~)BalWA+D?QoHf5*hG%n;TVT1L}o0*AU$KZi0D#64g0*HpnxS0F>#n(e^`tZU` z<`_o=T+5LxW@`29K(2LnAbD_O;X}uix_w&k> z(}^GjpM43!n=+*?TX5gUooeK-fPKC9d6Er1-k3x;PA3Q+YZlJYKwbTVM4yX6X;Do^ zZ4M8*xQA>wd!D-2VEFDuk12%N5oNJ`BtWtwR?B~+$Vn$ z(>s`e@q3xqKjOzZFtJonH61B(gH3uCcfHjTTk)uy*p7j1Q->BiUc<2G;BN=LbY!!x zi$hA$23ldG&4tBl+PpSrtf5;qRxv6{{>rJuU{ds#TV2pxG~mCrKZ6bU%&KW~C`=U!?oPXO@k+ z=qC>^w71t_qnq(L_q*eUR>5p<6B>eKbJPYam&QnPNmt*`eIfA7w(YRKvchXVOfC+L z_}~s-)iKi3QB9|__2P^b@2;Kot1u+`yNp3R8 zSjoKYf}dz*^V@0stTrtyEChk!X5EI~`1t%l^U zq%V2XVQZDT$6}i9t1WycY&eeLS#EeD2hwwtpQ9z^=07&Cd0`7cz+Ynr{qb;6uVn5O zy2U<}>BGcO6JHkr^`bn@uzQv&V?0_hB75>zIuD7iH|#c1^Vkt+tS`NW7iX6F<^7r6 zmU6ATle#JXpw0055)-U6@5YvW6ZqZd>_4M<6q=+0i3jw1UY^Lzg#4{i=zTsBzF3zQ z(PdU*`|sh30iZ&BW(^!|>9Kdo#P#j#ttx)Yu|6EbtbfLHVD?D_TgoPK?9<&Y$@k8R zZaSW`S`05dT5e!2wDAV{CDJ^~nvT?sd-mo?^<>6**+c!4V~rup+Y*I|cwM;fHG>R+I)yVRX%b6) z@zQnfD|ZSn{B*3p_AiBnEzHlt2^*7K3IS4M@^x`s14rsEKdoa#n+p%4P~mMn3Rl^j z)bbrsYRJ%uNn>Fi7}T~6hkU8?Cy)2>O{OZZNND+P7y=ipJEPgO@+`kgpeCK*3s|kB zt+-3oqkqeKyDwz@XA$gBBGk_uifR;Uhfo9F6`X1IwaiE|s7GiC1Es)j7Ryn9HJ?IL zH+6Wm>5u%Q09+K6D8@3sS8s-6l+WLAXalWKkU%iS`U~>A_2je1BGvAM-yLBTVN?QX=(p`9|J)!J3Bk(?e7Sh8!daI zPC*v_Fr={ZL`uOh{pk`-G#nlbS5Qe3t;8(?jl40V*JLA*VJ9nXJLiA-`1pYLp}e~{ z;t-d21gz3?#p>DRK$boXBm)f(HC9NX&dkllZKQ7wx_jwLP>Ci7@^@JH#_YdyywcWv z0W5QOTud3eb&$-1A4>oG-WI|J{%l3A?Vg=^jMYkQvR(!-=}f&wAjD13%Mu?$SbYZb7ItTY-qjFy=Sl> z?wh&f4Ir=u*@Y?++&v#NSu9o8q-i;~CQQO6m!O$MAEMeFuZCRHN|)HOVVAGvCX&a> z$Sr@V52WYpu6Zc)0|o&<^jQKAW#ax<2k>DTAnUQPxMQW3Mysu7*+2h1xDF@r!?;0% zxO0IIMh(#@uWx;lu&e)&Q-cwpeL*EcB5F{v;C$290*}}##h7Ql%@m12B}ekqb93Di zbZdw#Tz1}o@Ss)zO+d205~sL)gV=QP_-9q+E#srcm*)EJAKt@baHV_GBL~Z28jxRM zVd3^NPw%IJ2{`vIF4U+2O*AlD=I9oR-XQqWUU&>kwczJ%QU+3rfQvM_79WHm#bca> z#(}JIv4_{i!!$7#B%4n@?qczatU|~c-ijv23M(rjYK+Q2MAn&1==G~W8BC; zt-Re6D-Rw$&PW+E?Iairtr+KjqzNJ`nU2deS*1$v3<{o9E z{`e-^b?Ku~srhepN~86nn#sNBsrbe|I@EWjWI3f8_7i8{%QOxWN7rjb`DjFa#S``+K5^o2;oUysDJ9f1xxsU>UalHNU^htKuH_=PRt1pZO>x^l_Z zQ{P-wpREh0h(g3oP#jD0B=Jz7AGYv{cXnGyZKS^r(V%_D?_)A>DI4LCMqHbfXe?|2 z?d(=^wMLd=V4+Z?6p5w()^MzYt>rcu-4?U7HOQQDn`s@@<(hUp^xQdN1x?+TW~}=) zD#f->8=hTn+fYY~koxmD-n}rz(AEJH$~u#VNJb`V)IQ|11;>U<>&ncMJzZ+M!;~F3 zz};{{OQw4y#qHDoUaqx!2<0N=cMc1A*ZvdNxeEO`SY|=A~j*N|fl1BOoJF1Vknz%X zlXk|_$@x=~eQcN7Bbx&gSsP(oN1kt_Bc02Sot{o(VB?ewy6>-ch((c}Uc?eT?*?fk zyfP?}{y5PQeMw%Ld+M~{2l*LXNH9F3zo;&@bWMpmwB*t~Kj)w6=rNy0^p^yRvN0L3 zcu8*5X1GI|GYguc-<`3uK@_e+q9o$ICMh*z@;R?l5m?OQ`(h6p0uJ9fxXh67(pz}(dkAcsP0J$$6PxV?%ng3nQ-;z~i5jSUF z#v-cw8wMGlgd7=ce==?W^OxkMXX1`o#ocB$jI;u`;*(!vhfu8D%C6pD{_h{!NDxXfe*pZCFe6-JAXrHY``2cc=tBmP{nL^tw*ON%b4@A;JLhEUTIHh#y(CxlDJ&j?**94=Qi-z- zdv}wUGSI?sy=vp*Hvd_-92S3G{?<_MnlUfyy;DcV?8NA!HjYDOWrO}tn z4IN4yR=Z|$NcJA7P2vL@)Ocg`*{OI-IVv9GChb>|QB}MQ%Tyki@>ZMt@Epp9Jwfjt zYI`-#`U6hteQNu~o_wKwe9W04<(V)w3rTm`h~U0HWz|S7(7WHgJH5hT^{X9lxzqiwsrEOr zFESqPq~G=&fML|{DLU9%X4SqG9|ye4oU}h8KkJXjsrLvIeep~8Lp92_zre!|d4fyX(6_JpTxD#cN zv_C!GVSl24v{gY0*A?l4(N%~+c8-j(5zn+x&R&se5r{~a%v>|2d^n+jvb4#nYIzna zB(9rn#BUfIRKA{sNrPMM&^Go;@~+O$!#SIK1JAXXL#CU(_q1(nm=a?8(r1%YDnD58 zGBcxCcx?8?;-jXJK_9zA!H`uRuBs$3k>Y6WyFZ(gg^Jup^{iZ6xC0M$*xhYnjc!sj zbRj#*Arb1|UGA=qGBY!QHAHe*T3Tj+6X>LZ=PC^|Qb7{osL9$oI$$IuIF7|!+Zog0 zw{LtFRx>hpAzFeL;M1+&zXhdH)yAx>t^a%q!9Z(9^I9iOpqytF7Z>N@;bCXTqM3Ky zUL#g|-T>Tg;I5HD2c!x&H?6!kBO;J*O~HYunhXL0#F@Km7g7YtQ&Us$CmtjC-`x<= z$At~LXHgdyd((f+f2^(1)xGw9i;jAGagoqK^-?pZ&H0h93l_vHhDW{Be$GRUy}Yq^ zVbw*0R>Ez9GuUXK`qWuE9U+1rnh*K{{ZKM)j-iVNZM%bGKtB)-qo8;`xVg!RpT4bZ z-%O&(xDgRa!U=(4#deSP3UT+?%$~lZG2pk zc@}Dd!a6yT5Te11HcV^8p>CKv%WUcTD^^!@ASpNoF1k!hay>tt-1Xo&E=l4F%a+GX;)_llo*2({j&mzuK7 zH|dFPwJ1DoCN?@Af6)@Q*dI%Dt1f22&C0_fNwl)AB=gdTP(l0)a-=7RUg0(M-)Kx@Ymbe{7G3jn;rH!pv>|bD1#(E;*(xl!-9S~(2+Wc@fs9|n$-Jq}8 zIrHmMf5oo7fo0!Dbm)||3PvbF*N6%&{!5GHpsI@Dby~Z!;JsF;7?*KdQiPdJm2dW% zNmLSiO0d;`Kh{GPlbQU&_G`gD^es*6eNe*wNgmE_gD zb=UoG?tkB56=OnvDsmV;?F8+%Q3^i;CXhHUcPGK*zbA6UbQ%5P`ugPP=<)~{6|iZ= z=clKMB$4{uF|lDg- z_Q95x8)=-zK&}=It)=51)7nWPye|;}Xt0x!7f_VYDYKxH}!LOcV7m}q%&GI-G#V$7>UJ& z=5Mx%6y#cxa@Y!eAE-YR69NfHfkf5J_`DBa3^mh=TQz6mE&tWxqZ>DX6bot#&kYIv z&WNv7grA&t_3lv1H0AF=A{7&148sdOb%C{Ky0b{Tiu8fzz#=WMeBQmU$#;6Y#Rr-%(9WWU~_9~6^v5CtDegKTI&oNW(2?_<$tuXcz8-t=WR_Rh9 z@xEthPJcc&sl}sPCgqKrcKE%mVIVr_SUBEG&lYr9d%OBAkrfw%aZ`JnqrvRijN+`P zb>caTR)I8ldR`s*)vjnh1z|2e5up){w}aa7O|e#W72%S2OO@?@ z`|lOBq$M&x%3|Ht)JmleV4)Giu~c_+QM~23@h|sja--pG_*rh|Q$|#_j~vZ#r#5Ou zrY?b~qtcD?Vn!&f>OtWR(^5W$l{I8RY`2xbqs%_1wkFV!^HvwNILsvg+pk)B)Xv;x z(zG+n@7|eK^r>3*jCz;#fiS-p^S;be$#V{sJ_~xOb@rLU|&! za61L zuOailWxGt9b604rVK^&7UsK|4nD8wd_S0phT#+ezl386iDFlymyNST5QfX%naHE6J763 zqrpgHU%D*99k5}cCozeM{XgfbjQ?%(Mu0oL6G!jW(UO}o*&DF^nXqCKwX`az?K?iu zs#BtYsKzQPDq#m~TwJ-l)(YudrYstt(Z3+1!F$~kU_q zz(ohTp68p`ko1$zC(Au%Q8RUa;ekSp2%FEXX0Z%pO2M!E6mMT7-}!{Al*jacX5(y9 zmwf&3d4Jo9WP$}(o9RW*CfnZym zMwCk|$#>NFjz7N(e|Z~0vqa$UefP+?bUlcAn1~wMoLMlKA|?K`9U^t6SlZ$&NQZ!LM2}Dn5?hb-(oLQ z5jAt8=79nVg2H88iJxU6oRgJnrCS@|Gt;tJtR{xnO34&_A|B;P=tbp zhDRrP-_-SR)OA1Fb$>G`^>1S42WC~MU`KoVZpr=AgEUAW5Fg^AtE;P_feROhyT5+@ zI-Zmh@Y*Ps{D{U^@3uzk1^s)zV`ppoZ^9yi#w;lDMZ}(?z56p&q>2sWeGfF1>6GBo zij(70FTY@Cmz9x0#i5K?Kw=H0VE|~bv6mA{N{Jhm+qS*vCA!D+O%XNm{Li0IMoI5o z{6Jp0L^y*sDrbpl-!63`L?L1@6@jdgUJEx<61%CumvF5N3_jef2~QZF5QGm`u% zkJM~#5zF#4?`lp|(RPk0ISlHFa(nvc*Q5?v=W?@b__iq}0=Bw!yF3QT*6i;N**Srj zgZBBG7F5nWzoNg>m)Q~zo;rPY`$DqyU6OYyUV%MrPq$?T!Iw5ceQV;=>qSCKZ>e~E z`+ofu>tmxfIQ#v6Y(`Yo;;l(Re)4ylsDaW&0j;_z`)sPxOxA5r5;(r`B7SE5zVhpF zt5JZ`3(9t$R>%Mpd){$VOjI8n*sR;T=c=xGFb>Q8LV1FMfHRqLIr=Hi--Q2(b2+4S5JBwCwu;( zYU{m5b}r1QO28pRPcwD(9o#TGiUy;~?nS9*rE*fCo4dKy!Rwqs1~4;IW){EjZN6g` zVX8XYXsZ$)SwBjg&h~9L7rEin`SO8g$BZ2LtAsXEfs3I}KW5xqv0W$(BtKZsjgZ;2 z!Z-@1bmgWG)Ly7c490Dj@k<)kOmfpc%~~f*T*cSm{byUz!`jmW-P0esr?scMkf&3; zr_;K;2W)hbQsU$LXe2lyt}Owynu`|f3okzh2Zw(Tb(2+9Q~(zcCg^b>i3V%vuFKQy zqQU7W0+!~gO~p?C-fSh;n6&UA4n-s~2)5kngGFAinI{P?fiBVwr(|bi2_Nt6{cH5v zEFg|!(^Q1n^;0|1N3cUx%94{B|egaVI2YW9{*1 zE!Tg)Olb9;^-Srj*w|ROF(eub(woC2M{zO=$D6Rp0qfp7ZYs;=8*OMD^+S`G3-(!2zDar1F;YFpM5cCQC~ z7W7#89;RqX(0-|~SE?y-4{MRJc+IA9{<8bJ{83%tjp3(5E6y-NyPU?(@?wuRNBgV^ z0|%qVw?o`lA7dDC-_l7=Ru876nWEMoNjAG&Gr08atiN1AZAkm7_cOGr4Y$t@;T_p; z&+xw%-Vmc6!E2P;R%-1Raf0n z`B*-nqTA3)yl!`K==JIv0R&yIIn1O>Q@t^pG-_)DFANgaGd@YGVK$9gA z7W*|h5|YC)10QA~R$K4p3hIYFV;=&7An@1Jd*=ni;E@=ve5>aa@;%{G0#n)X3~P7~ z$Q_Z4Z=BTS595CfxcJY6TsGWZ9(0|Ju-A;7sx)ZTm~{-&g?z@f1QJklQJIeBs+L-P z4uW3;>c|&obAPr5g`DSk>?Wu!e(&UktoHU6HaBMg0;l!q>16KDML0x%R@#oHc?+)0 zRu~|pK@p%7$QE*QU2gXH_r^Ez^AR5JUARIbCFuS0=g-MVNekk($Fsww%E68QCLtf2 zE**9Im$vJ+I9`?`s7J_Q>PsZ{xUGG$td1Q(ab)Dc)SJ!Mp5J>=)3u5qsh{5Va2V(o z2twk0x+sx~><7U@E(+-+{`3(+j>1uKke~_`h*S>3(nBac0XiN5l;v%yDeTXm#&iAZ z_d&J4qxRU+^+uy_-#Ni8n@${4>DCk}%frY<^?c&Gt7rbyDO{Gboi#cS!wL1$ zi=Oqt|F}sL#*lHwRkt^Ij>gwCJ*;@)h+R^Abr@vG;C{)#zU^IHd$>0?KW+Gxk7k@j zLFQL{$uq?)F;U4JS_XAqqvvwsY|k!{Gc`WG+WyQ#kP}gOLzy1;gzC!6$MyTL<=4f< z#n{-Gv)$0h2!eta>+F7Xhk`@Sv)BH0*$;jMU)AB#h~C`Yl@u5M8`zDu*7J(_8hK}D z2NAI}e;E~%zeh+=;Upv^PIHx@k;A2?lb%Sd zpuoViQK*`zGl&aS;NJ&P%2BaVvJ#S#9@oS1j=@<>60i<$0Z%P0EftlxQOw8*EsMxN znzQmq6ADjn=*lF~=H@2MvD@kO{h-w2AFm%0XqSzSv-@}QwB;I zkxLfe*rC7|RyX*|mMjxEyRO)KY)-ibe>jI{896`9QqMxB)yUjF3Vz7gE)82I)oXjU z<%{9azlZiF$RK5ydOTB(gvveN(c`_!O_}m0anh!^hgml7GvgHO7qok79=RUQo4XEf zg7&LL8XW@==kUP^%z3IHmO6osT2G-4xzePsg?lM;(7tBYU;$no)kxzHQ7Fae zj3semooYs4%iYigg{l6iv!FDY0oBXWPp;{=hT$r&gm_Ta=1^yQbjFCd3CxVY@-{2q zNouyAZda&_RV{pbURA4YsIEX+`)uUr4J(6hqK;9Lt+oAuiiCn@$8}uj_R|Q~>>-es_;e2apeg>RPc!)JBO-GrvIMc2CZpx7GF* z1Pj>HEZVzmKt$>30dj(vmiqb&M6Q++C(3hrf9W-CS;x6S_wMxAQqV=tfjtYpepp4B z>P??hjYv8AA=E9uPwERp;oF^<+i8wdqF-OvRP7 zeZJke3j=jEXP3m-8aXu)=VHe=ZKTl}q-qLk1slr8eP0SBvA|{6%c7OTf<3I? zB5-e_d~rj~9Je84YLbphTwyU5$+{=jA9m&re|ipFM6YQY{8g0#jou6Q72c!ffoUk! zxe1H{e=#uaDj~u?S|ejR|N4}Rwy^Dz?ts^GF{&G0x3GhC_&`fK;DcUcNaEo4E?ZJL zhSY`hnpgvOkuJVfgX#B)=sR)Vx`cJ%gTCQ3pZ+T!R&@+(iWPOHsdftQ@(OYe4S}rG zx=}-FXZDK6Ed@5mw+Uj*CG8`|%W}J939;&fl8m`Q4f1OnI=FPIQG^L;p8yrqbLG^1 z#d6{LX#_Kiq=i*V*%WKv1O0gf~~mD*=o34kBkHz1+VI-+5atK8?tZ z@9Og1P`3KHUhGa%P*A-4#dUp6ql9O^T4#NzG@S7UGMvF7@p!Y9NGXWqjf;s%$!mqo z@(N*M$^Z8C@BO~L3=a>_FI^_;+3yO5H>vo@o|j`nI!<5pUook(Lo~y4&(~n(UgA&toUldYvN!QvShd zm*F;1o0)%#LOxyRd@YSEHooDgk`T1eK+x1OXK<3exB$@W8(4<6f8wtEcZw-{;;?ETmg|06ap zo+x7!6TO#KKm~=a$zvGB-dUm;SM0N8UL@Z!)v%21wqrAMdZFe zQ~&aDE}2ywe0XQar=rEZt*iPIrTCYDmcb;7Pk^nVidW$>An?Hdq(b`+8(8TI2?54w z#C%Pf-9vx01-SzPk2`CfK>-i9m;bs~jOf%PB)|9e{`O#T!%1<0wfXti-3R`c`}qu^ zp`oWH(LV54pLyDJl4rlyN`e$+)3U zjV?=o*}oC|NJzl+tCPRyJ16Iu7ad@DYHBJXkq_posJz2k6zz8(5x4x~b;whbfA%Rt z60bvUz^1RW?1kew3sMeP{=jxu{_c)Qn*W-;6ofxp_Y+I8a72b%gI1CKhKI< zS^HOj*xdm-(WlGJXzGugH8o22X;;tY&nF>W)`{|GI`PwHS9rIdYG$}7Ua-Wqa5y1wBh!T z8@7S~qet!UB}1->fTwOG1 zg_^c8=V~wH%6bPs2q8y33%TAP+Lc!pA=8@V^;jvKxAc5@aGz9bt@z^?;OmBgdmtN0@^v$51Lx-%C`ALS@n;=Kv9gusKXQ`(aKA#8GOVk&XLsDm%xNX2MYA5OYa zZDspm7<=C~{5eB9;;=(Id|uR}^Lm9A(i zlZey?>;F6@SPhvD@wsuf(1I}Ebw9RA;X|uGG&SAa-X_sWNlLGFU#4QaD~pJT3^`90 zpjddf`Th@4Zxs~>ux;%&4#9$JAi>?;-QC^Yo!}5GSc1E|yL)h_an~S?yGw4f_dfUM zfdMbTV62++n^mi7olz)e^H~AE_F8T0e24!0n~2|gr!SOLBKrHyBqzA$)^de|nN5ex z%*=cl7!O+g-*`TM;{sri!cC4&PQv)w%@t1)ICcuxY~7u25a71w=kEgZL;<_<>1=tP zcSA7*aPPzcuUB2R0&psQ;9B-H$FdTz(*b{IXkcSwFSfZJejL|)oF9F(y3Oep3cjpN z7b|4xsHj};FX;tePl_hpt{VEcB*=BgKghScZyyKo*n4?#PxujA@NPN&Z`+s%!Y!%u zDAFIk;GedkM#t=p<0m#uYnb-v2cZ+I&aB&l=xR{KnW{Y6_|kFX{TN}A>uk4nhx(+p zB%&QR9=_9dNe4p_c4wA`QknsLt_OLCm>#k{`}%IL_g8Y< zJ+oD*Om?cJPboRYxeBbwvqkgQOgAaq6XBeFG@`Y95=bzUNp+8Nm}DC3y1~OJCnKj4cUyCxdsZtvAY;=M)iK`4q$U0&HRPp*QSh3Z^y;tLY zo)i|zSu)ODO(wE6-!LGY1|1#~tikR=%w;k3m?jJRJB!_Sw$m@Jg7dPVGB;@Y4u)4XqI#svg7&*^@e%oBm zZ0@&siTKRc0*{L6U3MJ$DWw6@U-z|a*bZ2g1=OF}R;!7U;*k}rdJ#RE z^NnF^I37ZmXp=$Y%2rn|x4Mp|+h3K5$5GDa=1>~i%mx%5=9x-2wqM|KVJmr~^# zaTm<@H^aeM?Ra!hQFzK%e`oBTX&> z^V?8~5xWnKXM|F#KA&=_)udZtael9ifT3lx)>sr3_4o7LiQvZp>HA$xP{|cm0y59r zb4TM_f}+45GPvh~iM<9#j||eF&*v4TZ^sK2l#1CQbpWaI2jfp)q5iR2PRZp6@N;oJ zf;oLl{Pcx*gyjFa2VjYBw=**{!8EkpBc!KCNO9%m^)Y>}U+sPRd1|5u|DEol*b;97 z8wDXZD?4d+HG&4N5)dm#i)sT}Y2eXpwp%LtWK&XY#TGDGB@y6gHKN<$WtA-HoAj!G zEh5&tUSxt84hu}*li)e21_C+r>DjI(GKG8?G9rSRb3_q9dM9Wh1|&3S zn^Sw?NMltJvLx_>)Dn6ufN9Q_wg>@4Q1e}ECbpjmcZ%lV~iw-0zA}&AV?O;lcK41!Ij_MP_Z=W6>F10$F8d(Jd1UNc6@_=f6 z9?roT@1yZgPHsYy*zZ>V6lKA9v)*bZAM6u!hTWZ=ov$?v`8!9G1e{|svRhqV2^LD9 zu|Qsz+kb?HQrWKxM;JtM04J#ecc&{DZ};Fhx6lC$Cj2i%gNWxm;eepAz^T;1kE`Lh zF?YucK6eYSlNp@EJZ@AL;hNQYPZYTL@qyQ;D-8@OQ4HxN|7`DNL-r^~Oye{cH`Sh$ zWl9y#^z2b34e13Uv-xB5nFV6xB84b zNyq!2lK1x9_BD|!%iQtP5dnVNxEsdR4??;s)(3iwbu27OtmgiILOuVgMp^h&6%;t} zOURwE+PQROi*DJ?HJIk0ZJxBu+0b5G(eHgh3AuL10uU;f%x>w_!N|iltJ1iS)Kn}! zmne>>e>$hCBiH?*7t*EW?{D*S)JCJl7#T#P1-sagcc`;&W^};N=BQOA1P+;PXFSF_ z!gA|8(aG}rm{Nc*zL$^dMcqP;FxN(K9BY)4x=r;GytA zGLvc-mZHx_?JvDoH+~%S`UI_3NdaIDhmFbddbQiU2a;~U{UTtsSMJ>e?R7>8*-_2f zUJj^dh=P7S13|1CaMAir(g3S}`X-2y+uf;dOGU^KdE3M;k79Z2UC3HnU}4^_YXH;- z4Jwyy%A2uAHWe<68!D`;~cq^T-r7I9rv1EkhG=~a%n8)U}nFSsD%Yy(>QvFKPM2v znq#Nw$|i1^CyFv+m9f+B7bTv|nkR9pSgS@|+qYMG)SV7;C-mspiku=M2SwkL2BJdc z?tfzD7Lh+N68BY8xMOD}Pt8{btZqWd1BLVOuyt8Wb{8A}sOf*xP~VwSENZu4)J6q% zeqUG|!LXjW!t){7#i?FiZ@Zu}0=erxknbDB=4z{{Q9Td=WRJ?#-K|)d_?ZZM`9DWI zt7qwatBHmU;@X*M*H~x_`v~@`Uz`I zgk*T~67ovVajPJL+qf>GD_l-COC z4FO9Vrb$Kh#1bcXz!|LpiJTOpaDc=*fpfWPPMrHOI}wwOx19*K)qJ&^Ns-{b=0?-{ zs>2QhW*cg%B{}nBSCHBv(0yAo zwoBCx8|`^flEVYOk2~S&E;($lLkks}%124GnpUyCT4Ho-{ho({= zgX432pO1JiHm7L16JwP>#0Zyx`ypRzmxM8Y!!phkDo`uEiOAa19Iau_g;qg93OIL+ zd9kZ>guh%rt(rY;fM6LaR(V`zv?X28A^dV20*#l!kY2xZT0UjITgT6 z(WUssFA?K>c>N$T;D{IjsRuifThMxm`83X9=QwA7c`1DpcmqHK{0+oX+7napx4Ol1 zLR=*#aHrwtzQ_-p3E^v8)mN}8(xggil#%s-l1_VW01A8)ZrUhKw~`WXCW@{m{Hpjs zP7%x22>He3v5*TcU12BoseIGMwur8?fM+u2;F}9#_&U@V@k1d*l!00)JA5?uNPb}O zP))oMC-otu94ZykK9nLV9;eUkdKW*gkdECh8~ZvZ9}s!V03LB>~1jZ4D#a_((1t=yHR>$w&!Y+=gfdg7GYQ1Z^r$yBq~td|p52jU^pq zLRKhJ5j(he8g=gm9hu6XH=>Q9fB&P_1mGmfgj4GRw)l_f+&2f4L4~fauHnc8i}~VF zv9Ym^Es*k60JWIkzljBWTT4nPz2m~ehf9?UR8&;p`ykQ55%hGuE#baNF_p=9v(qaO zuDiSY@-grl4E2%+Zg{=M2NP*w61cEWNcmDv;=agk8W&CP-8YHEqd zxynoX<0;CYKbxE`*Ex3GVV8qba60k;aqRzF650GdQ#pdftxwgb#6~boF;g*U^)CA| z;Bb1%dg~KS;$7kW`gjGF|DvLzR2fJWQ?9#{rHd8I96xYsmW;-&vzSc(@dNT_V$#Vq z@-mzxi_z_E9>xKk&JnQ}iTWvgqCjxxm#1$fs2-JssuY6ixAm=O{%Wh&w_(xz#05@k zsy2sJEcB8JNM|K<=vhe+fG%^*R)Oz->bPVefnD^3(*D&O2ijzms0?rDJ9ag9(7D!> z)$yvb;{L03#o|xbaZfawxrk(e*@%EMX79iEP}<&bti5w}4Q&aB&y8mS%RYWe-aAE& zKgM$X>Hzlk?Whhb=0Z0LL>sWt`gP2Yc8gw4TD88aH@}D!yXG6gf|l-nx)e}I>BOO= z5A04AHnqr=s;Gw2^*G-eG`;=QvpxU%>1EYs!Nwl{S*>qeve3ALlFEEmwTgC)XI|yY zmvu5?^<--;AC2y^idYo41X>`hkZn@A40!-=I6`XGH>f~{k`(yApcsDkvfAfw0<2Lr zxriSWFyhYcR1-az>b4stTGi8NL? zpvQDJerY>$zfZ3ILVv@9%?|=fh;$b5oRp%m4&|oJ#B#FmAkE zJw5+8*TuXAMb+s8{h|B10Kim8m-mt!fABJKMw$oz1n@$^!oe{;=ly=U?fiGhI5J`< zB;o>0^*7hoG&D58nUBZjU$*|e7o`|L(#jDtTe`RD+4^xl5C|ar@|rz04<4GZTQJ`S zl_3`V-SV;pK=K&bl5!}N1N`yIv-Mnh{#IS zH7Sw9y>K1lGAWD>>#&i^d!0o!k4E;InQWmt8TtcSY8gM8d2;mYd2Z)u{eiO4U5Y;R zrs7AZtg6)ks(6F~^~-@f^Nk zI5fbsmB|!cjeGFW6!h02HOuHjlC9iYRy^_w&hY5$=f!z1+Q8H+_*`!465oYl{VrHV zLJnh;)*u(qBX{6IEhIE9XTPnZ8J`hhEqnGoK8y2C6Js^PhOuZ)FG{9i+9IX0 z3#Vv7RTp5w&_+#I>`4P98Hr`paI!SdurzZf!5p+eh1^8mwXG*b3#6<}fgd!sfdy)f z-{=^5AJ6AOKwr1>LIU;qB%u5wu&n&7Wo7&aUAD4f*6v_Yc(V-X0)Ra~WFQ2-4vp^h zr9~}y9P^@Y<7z>}ED(XnPry0F$}ysKs^1;&H?#qD7c4x=@<@+fRz`<4raGh=pW`e zR!;0_?D=sj ze-K%92x?Z5vmuvpfZU*}H0Dy{{e)8Ar9|#^{&i(|5JoGyk7D`J_iSiQO1?g;f!QTn zNT%BnfbSmkFA|IPV;Cui{c0wk7dq~g`rapaBq_cZ@IQtfA{$@JX}(nXa>Gs6OhsjO zB#sn}EY3Qc=Ar=nB*>ZB6@0_d6vQCgx3{+-@8BOlHvJwp1jhqO_`Ppa`Mg+iAurkB zSDS3%5|ONgyPp4<0h$l^ft(KOt!_uNncS`bN)b`fpoTxlshewS9pJ>`@9z(`sAt^_ z55@m82v{I4Wm;MqtHlIBNx*p*T4FqBU)Ws=z<`5|J(a;(<9ayt?c2A0HV5==+7shn?gX%n>p~`bD_;pV~#zjaYRMU%QSi zFzcwFuX72Wg(W4-rA8a-l0mAMo&6)M{LvY(Eu!19q-^N@9RT22bCG*hyl>Jn#tgh$Pe) z(a^Paj|+(jxt^@*ZNK!V)gYL1O8*#f;t$QB9=uETC`}Ac8VVc7)^Lqzjj`+!48^`d zZe*j*CS%)e5Q90#?|5ZQwmw|n&?M71#d$6 zQ51x1)7hp{+j|j?7gYv~rF2Z^wR|*#`B(!cDT#CtGocHtDf#)aHl$xqbG;LN8P6!X zf=BGdboMTpck1o(mWk*`Fct5l3ez&EXc1-b1|;lvO+5*aB_`LB)qo%%63>QDj84c9 zW?d8gu(e*kW(P~R2XF=z-VX8$Rtz~))SstoOcdwnEUr+}Dy>SP6IZ_6wa3PMWmRlG z)gG@o6u%j3Ve8_uAE=x%OSD)dx_PTo^t+qmeYLTr;#%^euF2ZKn;nAkpC&5cT1^Jm zsvubP9K<}@^m?0mUN$a1j*I>wO6q}VUKs$-)q6;(1cZdAVCFbE(KRsmRnyki)?hgu z=8B4h1pOVRuox$Qu~J)<@{`izIDkh`@Za;@iGhIuy-uT$^GhN>&G_(g-W;Solm(mZ zqTTcD5dnvtkfH&3z_|np4Ge^^2hDqM_<#GKUI4it(kq`4=5?6HY6&!DQWaCs(!w2D zs5c*HG8|Z5t}`75#|&F&V5)htek-Ok-G|Q7zW%_ex+P|D@EA7e_w)7qnbB*tmFK3@ zr#`pfHuhlUSwZ?jlr&AG^F$<5p=e^lhwVe{;{IHs<+1GWiuIW}aQRhkI*kEQXij~T z!B(|cKMGCAR3KSiN$1_buSzNNFw!;mw*|I-Goy+#Nfp0exljgIGI;+W6JOk0;X#i6|tiS6e zQ>(dqKbSyrhO~UUgahO zL=D{N=6^#-O4|dp{A&S~jxBbJJ=Q|2T>C)*nQXs*=OR46nHKDEKhXeaw2#1cf6O)5 z!WSw~b9QFd3z#dGhtLZ0^z-$_X3$&i@_7Jr^kgbClzQ)PDATWS-_gEA+mt;={pW7J zmzUR9iYXwfeShsZdZ0m&c zhx*?Rrw8z_jLMBdCE!5H!d0C`#9^@8?8Gxg;!e%>_V)gmCT9Bg1aSLr^riF_D){Ah z=4?RXVd2WA>LL9FtDValktlRvvf#Pwv{;1zHqq?)`j8gN@1oI0DAmm@KI}FU7uiqH zQvWP{#QHTP^%q3|d+b=nbw=UuBkX_=gZQB(ircUsf; zv&D_LpBNSlFkwTllm#FckE5pT;9+_IJvNIX8P0bGt%baY%TZT9rWE*9V&0{-7X9fw z*+poweZAbs(wIytm)H5J&%pX9%uR>fncuJF&#?U(<|eW&hENBM!t9FWQ7&YglQiAC zT%o}g2QZp=Z7LMGghjsS2E?Y4IAu}*dWD~)#-&f~T+Rh(%A#t|%yAttx`rfP(Xo+j zPmSIzIBJ7~s>8W+$1DQ@en2#p#*yx79@|w>gga~Nq)#>)#oWp0Y~=q4W^9~C=?Ni) zUe0ghamq?DPB4S0ejepJ8d()I^Bej%yuF$efH7h^1;-WpXwU*XIEpJ@10|ttpZ7p* z?AIruaXA(|^%5P;4#>BUiUXUV8ls%<+w&c{Y%<@=X@h=;2fypVL`jZsq{8(E#4R{K z;bUPf^aT8y%HoBpRak^D=$1~P2y8GmHU`^mDVk6LU9NA9K^MG`s8kD<)mI3&ag8dS zplFnqqIxMHHi}M=#GH(bj1u(YHwp^aN{5p`59I)llovSU|5X`^kml%2e0ED^Q*4HKZ%oZrc9?aaf2s=6~gL? zm*o~UO#fCk1n?;k$FgDi<%|nnh|jKXJy>$eyp;ip7?qIKbbdZm62lc(>bkOv)j#DY z2@(1e3p8unftmy2HzTQiOp8WWOZ0hf=W_hstCjd;!s}Y9gZ@u=#~pElnY~dxpRP`Y z$dT%bY%m{2(=s%L+%OPfk&hNgwnw>;pZN7p4l^8HqA(v~fe+e(O7HOUD)K-mTxk5>Jr?japPwxw~kzu zpS8s}wHxX~G?Q$a>s5yBnxAt?waU^!zE#rYanK&MNpWyrC`*I%#F%lK5YMDXpCmjB zoT~7OT+vgnlP3DoMV`D4Yji)}egn!uLzwvL%p`h^=Xe!)HT-(>2q99h=~}&C@NA{( z;36n;X++oR#E(#a08TQ?*WPOUzxwlR;#H8A#$z%P%HnZ9n9B0m>4RC7e*FT$1`5(x zs?vpXik^bP^wAIh@#9A*A`V!0R)>cVU!QI|JkDI8fbdRNrCqu$4q(8ex-RN;oZW;% z4h;zjX>{E3ABx5UAh{WnaRCI^+udndSmJ-zi591%;7}nbe4hVEXfq7$Q$XrR)c>6Z zb{hg>s`-5GKkn7bAv0^!(b28fn zKq@Q(g0&h$K_&=@urbOUUzfd+xazKF`P8;YGaF-t72aMa$EtP6+WiJ&Nt^dI3(v~? zTg)#G(g96`zn#~P=k}~xd!a+rmUBj-d6n;tpEdnqv@u2XzqlG%GLG}qTz7v>!AdgJ zB&^I?-S~{4Y01>$QuM49Ce;RMudoxFnEYZHiu*7z**FTVlpCeChmBIl?qx4zC$b3k zU*o!n&68kn3)z4fS4M0ON>R@Z^TCpeSoYz0~r1VL`!*-q@N7pTxeERDXO zc{2!55{krs+pHYLd_}wwy0#Xp{w=?g@;^vsDU48YgEL6s%6Pixt&%QRrKb1_RkK3l zrWPuswX$e5%vmx|&no`sf%M*~ymn~cJPal&bn~y~DEV+P-y@h_=l8gu$wjaYmiB?% z;^(IAIk7=j$N367mnwkO&4E+wvag&p5vv@p(UkY7RsPv3ALHTTywgsE2>p)LTO$_CAy0Fa8$4_^c=%MQ^x&hC}zOE&->E z358)c3pwqj07J+9%j2G`e3*P{uvbB6)D{1`^e__J&5w{7N>(={9{R#KLlparY~qMg zHA)v7M-pExNa`m60I1`-+34$xL=1k{^%$+wgA*`t98w6!PZl8G7Yt=|K3}E^heC*p zg9GO#`=pNk6`dxP%h|-n<`|r}E1<0Azk`5zzrOJP!egZ%A^DM%*A5m>V zuC(C)HDE|+3SE8ZGDD4F50R+({#a7)&Wy|p;I3UGd#`O!rX>k!wmeURk^ws0 z%GyonVHF2IA_K2N?)P2LFk_=UEF<}Wo81nYUKn^St4@myojf+wiH9CRP{7Mha{_{< zvxVZdTmezaJ3_WR(F?R%@g6i)zu-QVy)Mc?6iM`YO9q{_?!($Vj01oe$``>7)vdTQ zZzMDVKE0Gcsb3Ta3Ryk%VAJjjm%G-hJ>HIUtt)9Lrq@DfhLTjw;ah^EDje>0p5(a? zZkv_{OuFb4tH~FSlz+u46Bn~QvBQ^%#fqjW-T9QFBw(8Xaz{ffTO+|5t*YC_?YLA; zIH2RrGN@#DW5!AgIi^{}XJU^ur`Q)YP5s{cs}Yj^K%@~)N5Y)NAstqogScbo7mXVV z?Vl=mSJfgZR0-3?lI&Aw{Ys1ul@fHqfqCr3o&pJ1(hVQ@V@%_2I``Gnowh zTrv#ka{fCsnZYTEdcM)|02X`ZjKa4U_!LINp#ij5d_`XOV>Yvq*pFRYfS9D@R+}qa zypfdMN_{Ab4qW_D?Vr!9&3>8}-~I-*wBHi&Tb)i^v8F_)w>$m4huX#EJxW_XiVrZE z7DMo8Y__D2w;h6gv(hXKFwk^}*r0!+`CE$nIkmKtr35QY8?;clY-^y#n>5G`>>x=T zUSL|f^o%r1m-y`_@S{NNTF1m7dIHVDHHf9=WnW?8 z;SC|Zpco{x<)GfVeEINd3GCpauQSh@{7=qD+GPEp?!`*5KwEfht;*N5bol}?z34j# z`?bkbRo*{I!*wSIQ0~xOwF^`*U6s;XI>Kj5TD?Upg~nC6Ks1gOD)#c)O0-+$Gh*mr z5NtV6lZ3=@QjG% zThwI8Pa}~28Id(=mlYQ4U11H36WCbh$>(4mr_;>cSKmYzIx#&}b4GAoxMiUba0NYb zD=s-Mb`}iZzez&RYxqbd;-Yj}cg)56#svR*BsFn5i{jz8fBzKCKC*o;6NsMbr7>SE zB*^xk%DvUsM(j0ra5x9MaE|Y_G(hINubK?*123diAL`|GT3Q;h-)-EQ>Rw?<3Dkjx zhDLHy(%|rLhaQ8GCq?c-N=gbbkDJw6v;9{%`_P<}TR7d{>Wsif284?zjPYESU$n8>?pT7&vWGx}yUq$}Aee?j6`v;Dc z_zMoyR|W=@4Meo0iMu8>TbTULejTo(U0>aUd?sRWa6s?+P^f&5^3Fcja;}nm^+BYht??B zkbQ&|GBz+jLEugg%MqrX;){#-vCk5L$QwgNiK}|+%WB%)$0S!@<@*}d$Oej5{&Uh> zRdcNY)bsXR40ga|9K3Se_jXSJQAgI`zy$s+jf*5t!o*#KaGDld*yqF~C;mxcp4&AQ z!b+vu>96UcUXv1~;k2Mo*{dbC@}r_zjJ?9mBhma$Q#MUuk_x}PkIv!-z5Bq_9jafo zQYhso-jA-(e#}JuLAphqRgLTAzVQ;EnNAP@xJz>Zf>wzV-C2E3oQ})rt&CpgxGj|1 z_?$q>nk1#8b38k+Xp(Pcl-;h0TQO6zTT@k*YWpsJa7Y@}HlMnyk=HCti`VKdfn@=l zDWT`1ZU&aeT{jse_v(oFM>x&U@Pz(t?}D79$4WeOJmvY`Dc|y^r{+cCUtBh2JmU4O zyrOurJBU`cI%Kc9f_gzSt#@Ey*mJl-o);CD{|z`($aj?y#@uKFX)Wzi3Ijxy{Yt(C z2SD~kxLs+qp8L2;yw&Yj8X9=y2?z9ar|qu(D)@$IwveMy(rEn{x2DOPuy|T7=qL35 zs^Ve{MOZ(%e~I}?rrSEI_iZgEo;wis^XtZgAYB*P6oLh`M8>BPT`R%z%$zgCmBC}H z&crinYWe^Uch5r*IQZZ0)Rb{%myEQdjL z%#4gX-#~LF^NnVEDuID9BDZ2?G6HcJ=_YhHXVQ{{2J0W`8KaE{8uq|ZMV=v)NB26I z+$5<&QE5Y}t-Fvk?5cH@iqW;}%bv8YkEpk6uJlpCCaB;Z?hglh- zdmjj3(~fS)e_f}hTM+Qp6;AWip8yT?+(u8lm^n|-kGra0l&LCKQwZItbyh>!`s*aV z*KStqoRC;rmYW=&*XB}jeo`OVa6cJBV8b&s#pKD%;Q6du|hIZWW!)4NmW5OJfMnErEa-C#N*5kG8>YaQ^aDI zBnb*k)yxTOu6f53XQd^y^%9iK>@oO*C*7R{8SrMTUht{AmEpViuvrH+O=Jj!_{~wJ zglitFz7s$OtCamFvjWkK5&~DDNUCdwT08kbHcx5e)tKo8=?=G@C z`UB=#CVe@iQrOgj??&bZr7P@$&qaAWSZ`N z+R~@E?b6LKoM*hzI6MA|9vj0wPvHqMQ1si9@n@ za0LbTUXjsrry*+8mVFee)-6puC9I?YW#$@0%6>`q%q32zi$1w!*oe%)k`5yc`U%yd zBB@SDFm=@5B+Jn;Rb`mo>}jZWtlKos$xJU6kUUkVvP#Zs_3_GZuS6b~CbE%WGh9Wj|pv`w4RR;_kfKJ+9=NxP4cWU9>APQP!F1Rn>E*i8C zOQsC3MEVy+ho=R+DLRCD(#Bw-Zl%+!SNVacieMii1%MH@qSIx6tS=NXC*b|*;}VRe zA^(roWBtm()_8{+omJr;y94mhg*ClV*tdV!6`rGW{JE zBU<}_`X+;axb*lo%jcZX#NSA+Hi_P2$lvgynH0)WC49>X!@Cl*&Bc-E2-D}q#rjyn ze}*`k{Rr?-Qm4huOQOJ&VHqEAsBCPvOQZi5)91c`w1L@0D)Tf*Waam9h$Gs4Y~$*I z>9$m5_k~?sH%KFQlA7>3M$X%8B(;}VWY2ocS9xUc^2vJ4v?dRA z)X-wkGwbV=Q(3XtUd_q$XWmVJanvy-tVYnvTcCt~WnO=;&o~9R)ez16 z#B#=tZ;yptEZ1MRlmehfF!N)5Gf5UMi*m1KQ=&rEM56^B;PQ%(#+c*%%)VCF=TR}F znb+ll7n+j|Ii?VaT1HFz&HnVQiidiQ2k9S!_->SV3Yz!WP&zaJqj2m*qf#V|JhyV)oU0Y)_SI0vjAy@&<+fwJ2YY z*hV*pib3!TYS)bRGP5d>?MJ12POYjg1!hL&7u7!#9Ab{`bb6d@!cKNHf}q8SLeZ); znzBl+<4kNJZu>u)TTz7O-eYoebHk7cI2_iw*w_*u96vmTnC<^}mUKgsM2Dme?l$Bz z>qLzBU#f>pO0g(wF6}%yUr@zuomU~L)bvDnkgw@1?bIz_b3(rS3RSNl32qsX51HZz zemby`A8FdVwd0m(ytIx|I9W&MH%27#gR7u!onADmu&5?n(<6&{bmesLMjr6=!ZOWfMrM zo+>0&zFYKJcNnOKvgDV^&ZT zN<~HGQEcyn!2Xy1p$UY-!^6R-O<+j;S$Fo)5+J|%YcX}D#>wr`XdD%<$2PB&q{W>M z@spL-U#Ep;+sbeFIhJ0MVKQ_P&~O|jdl+@Ktw6LVbJs;!tCl>kj z1_j?3NwXiG8uw?U3LPq?o^)yZnZ)zhQ1o>(3es$G-(mu4F|otM-g7A|69{xE&#}&Z zw@=}sc73|ule28vl%N(we>JKHp^|aa(*wGV&-i{}p=|o%h}Ho~RX}zZaS44EodVIO zDViy3p9-W0Wb|N~rWlfL7y4N<2_l`MVmXRMzKCp>$2p7f+f3_OnH!O(DR)nP5E2AG zOPH$sT+?Nx;;IN0v1o)N$sRwjyz=PV57eqZ7@Tt2ip7{HIb1Cc!O2|=E!K0z8Jv#SHLvGx zO*=lwdp0`<+dm%8);wPyuP)Gpg3oXaedA(dfpw`7?m$C zc@GKsg$a#-x!hER)L7R~F`GYYBx5C`Z8!-x48^5d-&oBjQmpZ)<7_C!oY9cAv-kSh zV6c7&`O;9tSA7mEA0ka+9GAJeq$Z79f!VjLuiTZhINBU}Y|*gW;Ro&+B5YADjR5+> zAU_e$Bf)$k4~zOKn|8Hmv(*bUwRHVE7rFg@B^?nG9Up>pMa?oNjv}5B$hxnYAizTRMMdIS~VHx7=w-H(jSK^q(P=M28GzmTd#E!E>II1 z5A)$vLa(WcaDXIR!~@Dp@3t*iNL+u%QwSH^dzV4nB_gW)kYhY5a!^$#-j+(_p=U2h zJoqdmRa#;3?m?YYQ>fyrws7QesK#Edh^K9#u2eqwtq|nWPGE%?uO>xlWt)3RUrmOXORH+&i?i(`xjf^+A13gmlHfeo)f>6>-Y8(LK~W-y zBIcloZ_Y{8BkhU40F?gDK@Rh{$cR?0Ro6E>Ov%VtQ(u2AW8Z4HbG<)aR#x`@{?5$I zOzax~NWXWqTdChKN-?lmuH6ZOf)jYX6b&1J73$FQ_xF#Bi!(Pjj~&o&h2z!h^1e$) zl4=$7*%r07w${>0ZE1N}X|M##xhS61GotSKa@|LidtC1c2v`nA6A=|@DKYbT=3P(4S`LL3XtbIwObBEuYz z-$*i3+8F&R<;1FdwT0^zXrtW3+g63T-1Gfz5ie@dZYw}+6O$B2zKD~(Li^O=42H%w zCK6#nL~W4%`!lI!{o%{wq;9kGWOA(g+I=1a8q|X$Jd*e=vItUw^|M;sHMhzXY!Y7N z?gVts38dH$>ZS1-kG*4?Xu!jOHVd<_Shof3O~mG(a&P?sG0JGElZ_Fim}(LxxklL= zO7S*lk0>?lB0K$i;eH3l%$W52q`C8PbnW8(Rv5+S;e4acbvQDKR-;BIDIrV$E++gqnE3q{;8y>s2BTp! zBm{aQkU5sgc;mSj8a{aRN@r_woD=%#@m%Y?BUnIuy%j^5h;d)g;d%FuspX`0eGrQ| zbPS2oEy;NZs;c=Po8aig8t$lpurK2rb|s}&X1~71!oXPOH>Q3jHjz~yDOqbXNDh%f zD|%?pCAWpzE-d)Vx_9)+yC>*=cud#^b<$zho;sb_+LM+;o?_o%8OJ@nS_sUeP4M3rw>GfT4fcmLE zg~xdUA|Cg=l{_;}KF9O@`RH~x%XpXxE?C&UNcw35#`Nk@bwndUN3CP8K&Wb3D6b!X z)fQAV3r!rpsniK8ZlCZake)oLIOn*#PW~R5_jpRx^<4ko@23I63R?2vCyC02Gg>^nAn;n%Wtn=2J7J2sdI%9b1z&kqg+3n53z5{gy2al1% z1+?T<_H)n3J9I)#ki3jdK9#;#9-$8udcAK7vyLG2?~D-*Qr7SN=}=r;d}d|_4Gm3M zSy@VKcz(!-z|`cD5`R^|_wn-Mxplo(Az` z-Gg}Cm+xQ)Lt}yO(8RbFj-DxOPmSgGv6#e$kmBLs;Gp0b%DVm&Jqy}F+X?SSrKF(P z?DV?HKwW?neD#B{dR}@0+nc(EhK8Ej_RVZCnpAduef?=?eu&#rnc7krCZ4K1Uz}d; z%Vh*}cGSo0G7DKLJ(ccPh$V7=MI7bV+Nl9b)lXXSZ}5F zROW2)RNjBi#8iS}xx2J_u+;h~q^{%k_A={jsh_Y8UB(hz z7)q;1U8W;e3P?QKfSO%1;%8Db9okfeg8NjSotNi0C)BKs1x-;B3k)Ro8Oq{v`jHu{ zYW*)Z6PnV=b_=EAEH7r!-zOGhCZ@<7=<+Q3ZsqI-VtxGWSm={hyIdJMUhO=%M?gSW zduks)WqM+Kj1K*&$``;&yU?aXDoNF7lT(++T=c6B5m

    &U)X(i<;J_S;HiuWbx|C z+-8=s{oNZ-p;~9R(C|7Zl;1+*F=NGJ20>m z1>{DeslF-v|7beLz&O9JYsZakH8v)0jEQaAHXGZvZJUj4yHR5&jrpd(|MPyHFZZ0Y z_gd>Zv*$iqYxj9hW4YnSl}2Qeqf+U2XP~BaUl8{#gpkc;_r}S?z41QcPoNR^DbfVMynn0=&Jjb zf$!f)_0Ml(zuA3_$`UqMdZ%V)u7rF}t+5j#^M|^WPX}5mm|vEYf!Da4V49pI*Uc-40mq8fW+h zd}m_$UXJCE6x*DR5I?GwwS6c{_a#l|4kitsHOTl_g!deUe%_qY6Vr3u`<5%b-ms(! zXG@Bis@^15?NtV-i^eaD>_CbL?ve7))B+_`KMDiqlydJ{TC?4tu z7jBIU3A0MEqxd*+RF?A!x$ypel5Qb4w8bn=k<~`bq#1%->Z${=?0(wSpU_W*TGopu zXB!ag4jU8KlzA3|%Fr~3(q}h#v5SJ8R`A)oVRm_}_pA3d_RZQ}*w4L4#t(`_Z=JFP zZkCDh@!{C-pZke&dn}Zc-x#{NoYseKPD*q6ltr-|!78)(eSPXvR~DEjbNROq9C>+U zNSmx_s8zl7ll&wAHTl|!eU?NdfmbR?hdRt{!Z(G;|osCF?H^1Y6_Z| zPo@+9G2_1xD%|GXs53bNwR(S~IM7BmHo02OrBM{Tp}t)rm5I}#sH8&4wI2=>E~Ft> zr4Zx^2j^FOyGT>1ctS2l5ltaibrc!TiWa$Ed6GJyBM2peaK+}A~XYQ}Zn3jssTffir^k&nkHx?@) zH(%bymrUEiS=fkzc}ZmmP!!=fcrd7hFjQrDvSfWrVRQ&c7#hwae6pU$wks4uPBZeb z2tBu+=L*ZK_l?eJZ)fh9<9i3IrH)G3rJ7&9o-3SgpSHi9W)Ld6WO=v%5@9A28mbW| zY-|F7;)GU`kKlC{6$-3|WF4*%DSjF_F_xNbsQJqszbg?0ug4_});q8dPHv(^+hr8k zQipeCxYd0XDOaS-AEX8j>voLJKvB{ zOBmD_>RgVtliLeT1DA6rH7~1nz6I|lI!R{?gfe*0amgapPU&!0I=R-as<+L3azX5QPhyrBp#-M0m~A=D@P6LOecsJ{zRkF8c^r*=_w{5o zIF2EfSB`F?{w(zrUH5A5b+?bD^>~{N?{h@a#l>82e6frL)umKpSgrSdS)!NRQGOSl zoWXm0SK6rB=6ukQ{V9{P{b$!KJ(#9v9 zpMPDtlGkaK`vIL2o&|506$K%|Q!AZhVjfP@Plp= z>jZzN9gG<#tM0C7GmB8c%g_HVQL{7VXNWyuOrx@5bwZ|8Dr=XGq9W}gp88l?e4Q9` zXH`;q2X~q3y`V8ZilO5;1^&^rp?jydj!EgN>!{Dlu6+IDgwH9(ZW;y&ni9d8Sze$& ze+9^)n1~UuTOys&d3u8Njeo^=G|9t?{Z%aB_&fYfd(*4uR=*(mGeZ_Vzp6yd;3Iv} z#r{~LUDu81mvImbIZMAxe0fhM1qo}q&Sv{Bn~fF*osPFtwVs#Gp4ZNv&)1&M72o%j z5kgPvzlT%XK^w#gXw1!ID|y06v`GSHn$(iMh)u?|^t;;$VsaUJoeCSD9UX6szMg*= zeY74{)P8$O6?6U4_3%6>$==~mMtg=Ii~V^{JdYuRCE3ThM8Tr>=gzcLi2}$aePUE$ zOv22NUP8QY!8~4;W~raVb<^VwKdee@aL12ssOdTsz`bcfT+i8pEh3sE0M#-YcabC- zCNW6qr6L9oQbSjW-Das?5YmKE-G|IN(18;ml%D*QN3tohUNkj5_01}kkaM5Vii6o0 zAkXR*yU7Tr*CYmb@`1r^+65S(n_*1@jaV}ZvJDvC9Y|HsmLRMP`>d8n1j&*}>&iMV zMH6tJLFsi`AVQ|+j|*60>|^ItH$_STZ|E?PRbu|Z3ZQZns}WhxC!c}lwCRVwKK>uG zxt)JkZ=<<$P9B$|-4?r=Eq@>B5uAQ%nv{N!b2I*mh8LOfn%jrI<1Q@S)cDg^(VPj7 z!|4F*6C4yXeIBrRmGiX&%ckdl6!LHy2tlw}Z=zPINHR)?goMOm)TP65uBJM#%I5WS zy*r+3-E^Gb|MF^+|M5c71oG)d0ao8bLg3*82B*c@XQL zo(aAVdzT8aD~_6;`^)XE>(DPX`NOgD+f`*PVkN2aRv;45sFYz{n15@Yzlf-T1Ri`X zCY04v0Tw=7>9*iJ{f4;W#0>38L?P?dAX|)_z!B-gJbDc6uL3fdYFQ&41|WQM9xmT` zbPq%TchD-uOsl>^*x0z>PW{>5+`9bL8q06Nmiq=QB&oNoU-b=ysO4Nw4FvDuZ$C{M zH4Tg}j#gFTdKFrcq!FR$0YI_rJldw*q7P?Rp%%dpAiP@j$6?khT2?=r2MJWsFeQ~K z+I^IEuIC+XHE6_2v?*VsjW63SKlf4iUZr5%&{RCBNXK~@GB9z-^+d$%vvW|_895Fr zCH9?}LmZl>XZf%oP*YkBfr(}avpXKTg6tcUc5vQXamAE2pQ!}L*3a!&sT@7;4;vDR z6e^Xf)zazAUwg>yuqPbOiI(MdIUpb_E8FRM@$r}c^Y2Q}$Gz{{i0|7*&-(_W@AHJ- z`>EPg2Afu=t4;WG(=`jd23R%luD7FOM(8TT+;MUD!^D6lkmpQ37JaUQz*#^UD(Vpy?5)YCylyB|pK1SsDuu!liRlzlMcMw`p zG@LX!Ked`5-qwb*_DN#Z`(%!Tl&(`9B4{c>)5@pxpFmF?am#!-IqW5O1~K0h1o9wa z$QZb!8btUqpo0U}=C)m)U$+TZnhmEyitQ>EFT;DNrn2gQDn11(~RrpdDBdpi?@^Jyu>A= z>udKmT>aLilKC@!CXGVUF&fPvOPemV<*c-f_amSKJS!YGV+8yCr`?#UN{l7@sD8522?HIYlDQ{slAdSnb;wYY*(k4s}4H8H_AZ zk$IToPzvLB{j$C|W$-<_QYZe(Xb^V^Pw459A8U|^!Q&$;+NKoy8vwHRFrJt8_22<* z7GCq~g*CP_Fz$)6G&C)Cs2C~`_W~4cBhTY4a!0(;HMh}9 z$-`JJ+aP2(e@ghY%}9=u8=+Q*t=5s5Zue!1<)w_>dVllX2*X2xl!;qj?T(P!k1EaP z3WI@=p64mM4!0{^b@ivuv2C}{%}A$<%G&M>4!fLJefbLI^66~ue?8i4xm2N0IQ)OJ zT#xDbn$4ChX47vkXSGrx6Z|6y!6y`2j!~FGm zj{}QRI?aBBz%7Yd$@J-Zmr^#1ifN->I1h_Kd$qx2;;V;bb9g-V2E$``)oOLU-=1y~ zDP%0Sx;z4*5KL!svdL%p4aQK9qo{5WS&ztO!rijf>+9|JM;f=E?Z>Og@kV4d?&LEW zR4h4s4yqOYsvY{Udh}%u@aMGbG+wWk56bjrJsij37AYlUl?=z#OsZ0iLYPobpeIX< zYbFCzO4*sh^W*^nu`1-5Z793q%2SUCe-b+zhnsduNV50{ZoBz~Csz-kXX8@iWi+hf z@6+Bjjc|3{GlB{9ILT1UodIBlu8!&E&V4MSP* z0tm5KLY)@twee)y$E$6Aqv0se%k~iz^O2%|=b~IkGr7LkJAG2Tjf`48LI1TDo&Q$H z3MHYekTa>6`})~%b?C61*6r+o4$b!KY^Cp;zRNVjcKZb~YJyB}#G< z%Dg3V5D(cs8YtH860&%bRoaM zh{@g)SsVnKdSYi{TAy~?DfOh|UB`W}*0#%pI!YLF;qPnI5ekw!F=59uQMijN!i9qq zBw%2Zyt8Yc>+l4lz_@Jxn*$XX5~Fcsw=XY-okp_H{u}Cj zZa>&<#l~mgV(GP;-S8sim%?EG-PI%hu{Aa55cyO9%nLre*}^;m{~T6}`J-H4Svl3|oCkiJ|1}gn{w*3W31~tDzCR&{i$meE zoI4or{=iau=Y4Bn5+8ZxoOWgwD-P)p2~2R%2n@y;^jWpne?GDx@`>=xd=Xx-Vvn;V|`Q{Me>65|$z#v}jbX6lpmR-$+C_p*IhUO58OK zI^2xbu6?=_5DZR5I4nOjnX*4Y8!-a)RSKG(=PnM_0(#w$k?UrTCX~jHn5iwKy=V$- zltWIWD9s6yELnM>By`5u^vD!b2qh3JDBo}*_XVd&e>yeWQ!)5QT+wUL=SdaDBk$d) z3WsXdcGvqCX1(RACj1_BgEl0=u>G`mdFgkFCiK}mlx+5V@OD}Yxwva1_H%YQ+e$Q> zF@d?mP&U6>ORnvHv3xjRW3yVWQm@||jVFyI;8wkTFpTW~g>!k|9}Jfw;CnNKMgFHX zf&cNp);LFG{S1p+C@P6S&_asI?kH0@n@+y6`J#r5P2Rg1u%zL*`X7uL(!-nyEF=jv zMZo=&I5v-wMPtFKq=c1t)p)^bk*TnRa8S8m$kI_N9isOPrG*1D*J@1{PRlUO@3f~~ zNv2#y5d<2?a_PVOki8O;de=wodsO!s0@$7U+4AT@;vY-v)((ImUDuqhcvCq#a}gE%{4Pf+ATiZ@8F1D3j65w z*Z0AhZ8T;p8{egWK2Is+PV4Uao){3Giq)=Z{ix{`f_S;!9RP=dr_yTqhqY=|>Nh$c z{0C-g^Ds@cU-68STq=*H02|CZ1)4RXU`Bv=!)DW!3i+6lX62T9Hbx+nzvfoJ@LheR$-yivJPy;RhAkwPl{Y$JdLOin{1i3PqeOSD>Z zyxS5HumioOxZ+$etq9`U!3L}e}{Hl8f}A4FO?ry$z>V#?O*)wg61g3ia+_i8tlsi_x68ROm?`OF`JCvUHQ*O{Vjo4om0;k6x4wAFSVx{;F;#h?iZ;ip8JieyHY|4?k{%$D+DhhyJL^6zmaZ z5SEfJ8#R=>4v%vdlh2nPOQ+jKTZp%Gqlb_g{koV}9VD*lxUw}D8x&FY#Rij|;Vu{3 zK_Pq+KTHr0P%_pj#0reM>WmJA0->X}j`aY*k(CT+;I5PSCZkT+n^;2AqJK*(`PfrO zENA0#fs8DF$O45bR6TES!pf6G7jiB^o;V$;oJ=G4Uv8B@C&`>^<#^t zdAP()BlRuR+!Pr>iRm8&?884s|EA=y;pQ>D4bVi`&~|c4{z_KTT6OHD*XG7^KcIOZ zvRdB(9p0ZO=T3RGO+rDA{XsA)LO9&6=jBSJl4SW!5E(wgX`Lv!P?ADOFF}O5ZzCfB!MflMu=n1 z`1sxk2oPc@355N?&jKNgLfq9+@0SIOcuMRp(HKgFfm0MFeerG0lo35~q$$PT+f%$e zLr1vKy(V~=QD7WXwnDIZq~Bd+J_tFOy1A~^kid9x?p_E| zdH?`H1p*^KQrb!G*2`3x$&nDOPCj3+)VDNJ7AOpck543#RoRP{X@rfsx=Ni+UU7Re zl1d`VivogX?4{D+I?>G$>g(m%{FCD&|1fa|LA{A$$WCvVy9F;>0lrsVZGVm6n6?@{ zc!d8uk9(cjOir;>s_k6CkXf#`s^a&96r$OsO0}6UkEHnz=Gwul;pz5E1>Bs!t)m>b ziq+omlkkXUg(${Im}*l`1zCq`S$LZkcxMmJxLMakG~z}Q*ALx>K!Et!l@c2OV94Xs z6fou@!5Xdb?$VI3^*O`C^9?)_X5}&U;ia*n0Ara7>;OYRyuX%R&}?-QC+PiV4=XtN z$EnvL+?nCJ;=cBA`{3fP(KPo1tfITP8xd-2a!u9)-!+G;`j8Cg_;U3_Oc@8N;=R=X z)O^`wbpR}C<;Z1uc@im!O{`ntE)uuZ2pVal0(2F$jlnS@*dgx9=xkUUT}ZxtClFfb z>wL<9L0}(>P7H{jPp|~GB04M7818jBRJI~yb$?3TENy^(%>o2AGmYUaT|XG}w(!_s z?rwR-=aL9z$#1cw;W2j+OoJqyC(CR!!3Z z&iiI>2!mF2u~My8xm=O4`)P0Mm*>CBrIN+jEnxoL@fdUtLu0{NHv#Qc`^6w?iTzrEZ;t$2yd1~D5H2rmEwTBvfR zY|}dsBG3b(C10-Rqvx@GQvJ$TieTD9N#JQ|Q1{2bt=Ab91gP??-V31=##aITw=!bR z;ClU&ppeI#F2no^Q~&u#LJNTiDM;cByn7MBw@s>e!iRphaZYI)!hq6(^dzk77Y0vSX-RX$$X(#aZU?P=P&31n@zEm!kp3ODAI!DmvVj#S8u zm1!w&JC;~&t$AC3h;tjy4v8L=$dUfX+ESpR>o^I>4FRQJ=C|)-B~?O39PgtNAOM+D zVHhDcXdWRQ#2&w^(Z@JaX~z#3F*hg7#=@GRO#yezn#iI{|bblMzu&b1C4>QCk`m$)SK)r}C-k zofiB7E|puBBU}{`gMqorT-rW9e?C~=kE@p5SoiQVY}bK*UQGe=#0oN6|F#)zo&80! z-tr0JGhxd4y~eND9w4dE<}pA%&g=Q$e70Qed@_H3x};U7-xmyrB}tP^)|0goBOsT_ zK813e!DeL&<9wL(pI2`Bw|J=Cruf2sLpj!rKH;3r5m<=M%&4S-d1imP&_$%=mvCn? zn&nBgyZQzQO2)NhnrC#9IRxf7HT!+42h|G_-vGuVkyU|vF>MsUczPQn*b})9IzM`9 z2~DOC5{iAP4o#~JHX>%`V7dam{V}EYH}D`bX>E;)1FyGp$lCE?SWp<|X9sA1&7Vu)cnxf>S55ca2N03fGwT*B824z^NAeKad-9^FB>fCf!MjwZW1 zGnepozKqn_-0kjp5Y2-L0$!bt8U1c{tjZy;-5}!Ev3d_ty{L+Z`Db(_VjP3VoQH=d zoZ|?>4ewn8L=hkh`y3ivA09ReBLO2!W0iG$QWrF@p$>#X|FUr9O~pfwk7FmpQN@yH z*79lf1h19Q@STKF-g{tdMx(o3ESE#G`HIPF6CN{V#&OLvuQpB*eO#~qdnBD^Qwyi{(m<5L`gX6ah0~2~_j3RO(F}oJ;=?^YQ*~TCNq^ zXZfIZv8G_r*?4PQ5esQ09LCxb5#ks60InftnrMNBAO$j?X%B~jdtgDtx?>`cYyh_i z&9WNme%dnZO>sqlNXQJ?)E2$L-K#N3wZTN%-it_j-8sn&5?k_VHqQqU<)l?q1*lnu zu;G`?+6jo}hHxf8L%>|fgAro_(~PFc5U+q}9H(3-q3aDGWv+zsl4o9zqsW?5IjLTgYUg7sYfS`-P5wTY+SkkFMclOTpW({y-GE+L zfL7QY@Cqdhv(IWQT39rlfe^|9!rG!x3TnD8@gf4IiUgCWqb{eEV_rw9D{z8aHDMSY zDC!%fZ5{y_Qcp-GYcXAmQ`nOO%rYhz9E7)o=55LZ?(e-EXl~3Et&j`olrfXk=f4}X)OMV z{u24%3IGIAH)TX2&q23bR|pdE*44x_mBvvK8#svtRf9r-LGQDoM_0V!(po|Giqm0i z%`Wd>;M`4%VO&Rd#>~ zQ66Tk!nRsojo=KFccCbwpQb{=6gy@_ol-IfdD zaUl{)OR{;H&@m_?Rtp;_B<7WI4_6nY2O#eJ6EPArx`Td&ilwuTA=U`p&L0&m6pJ50 zqd}U+-k5+DY1NiCsgv{9zSK-=TfJQLJ*ALxdb{lcIVB-JEHg&e&EsL|Y+TLh)+xmX zv*Loy4CXFbzL)f0gqh0bZdI+;T+w!#s;uqq@csNqczHvhIe~cD^1RAsx6%Asa`FF8 zTwuWHU)3|R6jKCNEDT|qiUS>57MyJpIhZy4tzU^{Zk>qFMG20jkmI_@PokYAyUjZE zfAEeCK-L8KC$h3Yp(bWB?;?sr9h}dS0y7e9+KyH91w?N)2y+N=44uG$8z*JzAoUnH zHtOS}J40A3D6YgOczcW&Y@$D*+o5|SYE!y!_k%`OqkEdwMle`RPb6{B%G8A!;jINH z-AA(baakLAy!S!YlG74tECj*_#j1|)Nsg*_N5BLOixI-X+1w+Df`ocYBwaCjaKNnf zP6Abri-75)H35!*{s8%RCerkeLZ>1*2CB{QdIWOT0A4hh7ri?FCBH(b1QK0ZF|V!GN5FMduD zei|=3QigCu{L7N!)UX@^xqRL~uXp-PCQ_<3n@^Ldm8NrewHj6kAYL}xou;=tuevw9 z-R&VV@V~Ssx9cB_q672p3C)XGXA<19SLN+@3di7Zp}=^Lvzc-ao_$H0I)zOmObx$0 z7+S;65gWAyf4*@^oG*S32y85uiO9MLfKF@}iIK^N@H{0{HY^f>F_Do{G zL7pl0S*DJ`yB7azIN@%o;(^1WMuKnDs7)3{qo0F{h^9)yfa+O_*)EnkDVagYL|Gd2 z4w<9>hd3~uYnSh(9JKj^t&WxZi@q#PAiRX6TP2`v;2uRnn4A^bzbuQx1;5USU2f8$M$vtA6H{9e%~E^2;)U7k zyMxQ+rM(@LNWy!um_0lmZ}a@G=*x-?>!($mGdRQ@849T{Z~c;*BE|9+n~jf$)*hRc zn%{?0nT)z!>J1*oU-I&Dck(6i#Q)3j(5#d+V~S0`n(}vuE(Fvn7o2xCn!vra(ejIm zf9NPw97dISFx{EoCWe^*X@|+q3E#tV9}c2Y>j>Gb!WB;7PzuioOE%<#O4DKOFElii zT3jcT*^h&eg(UUISP_P=EeX95kd|fv%$b6b+?rlR<1r$LU;YM(b{aWR-CI*oYUN zm9h9};JMSp+XOD*Xu<(v0ta&`qX2-v7;}FtlkpoQU_t)OPc98m zYc^YQY}yat*>ySG9nYcm9Dl9%?tklDbS2<4joDy> z!C_?By>^^1E}?)*$PvIf=1Som;9G~?#@5h~K6C+)9E(ND*IN%tH-IKL44$oeQ#5tZ zA~gaDvp|E$0CP%%tLVcWB>@4nB8v;MJj@HnrX{xb#?H`GZ0jGMSgP^Gh0$PyNf1}i z2*yg^fnqh=r$a#VXN#2grtepUoyJAqzLiIwv98oYEgH1G{ zUSq>-6Abp|$wxDz)>8={%0&q_exkXEh+`a{)A33HDJc}2JAcp1U!P!QS0n>lFc?__ zQS@sP_gqkfj z@=^Ynsdm<$upLzJyzIj@rdW^C^PC1z_ZN*UdC)p#^4>bPb6edXzYiz4XuXehP2?hJ zHFD3Y2@pScoi3KQZvCv3PG^o0n@2j)o~K(QWByw8|EY{rUh4uP>#_KgNls;sz^m2| zu7A}&%l+LurIiEunCyFj+El=*dI4*=ws~SfK6TnOZrrsHIS|_I;%Ud&PI=Y#0BCdd^GbrDQ?18*RlNgJ z5>VOc1opl_;tU~kP(~cZ!K`&rN|?X2NYp+M0E`^H%b|X;?7XbfP~h?1!Y(&06IV67 zoLr8qDW*Y+X6!ev1P}x84a(kF(<}RNr zZx3rm*V^uT*PMPzoQO;7YmohzC3@ZK5AqG|2 z3cyy1j@t-wvS&!L9_fbw!QuF>-A6+vQNAB=XnN-awY2#?U6xQ+$SID@<{AvB8X_^ZIoH6$lDU_~vaqS}*c|)jI z(Di-*=6=@d3HY7R&xpcJCSkC%_ zt3sMYc+Zri2o>}!zKb#_<42LGqEni9=$7!0&dE5~iS)6$A#T@1QvImI3m!hFn-{ST z9^ZSJy=(!re`kldBUGJlZ7ojT$SD*HKUEQ_O?!Nt1Dx94ZGg`lkbXLCj z2$#d9uY5u8tMXwOgwb>yv-rF_jN<-HDt$Sy=iAc_t-(+v`7Dds^yF9Or~bde2yEH zyLqz?<*y}qOTJ7gKD>jp`#3oR7YMNQCo$Sm5RPQl#ctN!jN&493=xi?Vkf-xW)sDo zw}7b)3{tzV?^_j3GoGHO(~>H|&uR50zS8Tm*Qh2gv&XX7Blz#!M!TnrW?Bz5(`(J$ zJG~_~p?Y}im&+9{zqxluVOZpHUU;PPr0l6-O+G)~!J**Wc0=$^m#ew$wz_D81WVA4 zxm->sGuX~x9r*uB{V)4#Uiz27Pw@;NV3W)uXu4J%=Jwyroo@bS7!gL=n>8M;%^J2| z7nQ-cK(RCQde@a!Gpd(=)oCnKH17cGWztg}&CRvsU0tCNapyT@vLg5@C z`&nTdaVSO4gmMhlVS<^3Cy#d4pSGJCq9970#v@{YJr*QON(8Lh?24fGYzYdFlBdiF zE*wPvNU`4q zC%TH}IQgi+Q#ih*GPrt}HvNHuQl=@keYqQ@p1U%IohZ=1$YUEDLZgBjSgKV0)Qw_@ zDgu>~0n+~7>)8;hJI~?akugE#n7(u2MT3cjH`$C_?S-OX)7SHeB3pE{&G!W?`intV zl(jgGdj{W};9{rLo4Q>_%p|v!%WeONjmT+rS-kE-R<<_zcFZ8Os`dO!sW07}b|NuG zOND&tN}UE;V$H9=_zYd9rx~ZSCQy%2HU9k=`z1VGpC7LnG|E(Z-EZH%d)e;|h9|s# zQd1Gln$6^7|8Ow;>IlvM(Ch>P_dEqnr7d`18~k?xu$fB+-(R=Wq;@7Uhv}*3iUdbC zF{w$qUp5AJin#t+%mgJ6CVk7R)5fvLyyZ?Pk(Mf+B-LmC3f)#A52CgrNkW{`*IJ&h z$TC{J3}*5W5LWI;%sY&N$v$q;852GH9sFsjCguVw_+)*sX$~CcxE=vc*ANeqSh{G>2HB;0!X$`?o<%u6M~z@aZ?y>!4jmbB@- z%gEY(+@kJ-FC8{L%J%E{$&npoFBT$73aRxj8|3Nr0X_%12H-aP$^s362ruQ zGfCT+YP;EVF}v?kxxqVkww_tJsab4$C3wXt`QBf7wW(d-;)#dP>!3RwGhjCDlm08X z&^R@l+s)PcwEXkq>a#|zcB{c;fA&Jrg?2bZ2bCB3-H|lS<=?y3 zZJ)C(teC7PuQS~>j$5SdA33cP+ePuM3!ell-*TtDdr{1LzNpma=N_Z)KJ4TZ$0 z_uL)9hWFs>v_C$@O;GvaKHLye&6P^yX%_uW^Pcccw^1yo;~}Mjp74s#YHXOEJ2r@@f@Y*M*0rQ6_hxQ9^2f>~S2<#BoH4;)#vTrE+Y68)v? z<2j$D!h#u0{1d^YM0s0+4HtGD6Nl-sW6*8B%rm>$IIyzF|7R7df{4;v*{bHz&yr3x zJr7umkS;vO=wK^FyjH>}r60FcCJtn#)7x2%9ueiEef`+#TV9#c;hYVIIQD>XPC^i2 zQ78UM&{Ao#EdhkxWG)0O2z{twDMwGCJTc^5wA* zoDX!VMv`PGY{JEDDfFDQFn$8_U?IeocIocq+EHIy>_KCyjpgr~^h(<*^|LDsGq*e`)BIPCuhs&%p094$ zc2;{m-(K~6euSe@*5F@?+jgjpr18JsEqrC;CPzpV^A4H4>)nBWcOJjpjC4Am?D0+W zeSIuvB6_mkTp74{GG&j27mJNO+`=karV%)|u)o&DeUJUCcD2!SxT zg6(1l6;-S`I&yW_BP7i5Q;mKxY-a{=v<@^~@S{dS7KC^OYk88Q&>s#&#GyPsoIRu^RzJde%3K5S#enZ z?Zmr$x>w;)t5~~!ziC`P`aZhtw0GW4=g&|-WO7j`7W;6q@t9#%Q?A$Zkw7kmpWE#i zX2j)Rqt$Brv0J5%IoQbCqI)fP_pEREW3=oTnr-#tl~OjayFxdZCi>>k2f=1w6R}Y z-clvzH=n4LXdn-n4W-2JAw0-#_DNRYktUB9l@hBk$Yf~$tww2blFn(;uEZ%^AKJKi zZlN@^|95PX$q=Vf9vL&VG`P})v;}}$@Bg9&1E*3-O7zixp(09^Qgkn=i_5* zt5?^^d2yyAY%G_G2AGawBx#tf35U)4Vu#C_5|o#64o4z6UXdR|Lz9`<29Sk z`g=5E_K!hnV+6Rjcx*S@a9B)#tV^aEQe*`;apyeXA5)>AE;WwvNxo%>0hhFshf(*e zkM1MM?!)R*AtvUcMnkHKa;daNDahgIot6%UnzbHe(z=^wc6PtZW`$-dX-FqgDzcRT zL0)76*YKMEQX7Wh&43+U z->F{>l1!PbV5cFH_o0C}F@hdOkgY?^ON}R0zw&YP_^_bO71}b3{}9Sh@)yMe@EEJ{ z{2F;NG_&v(%Ct*_n}ZZFa|z<7!xE@cBZ`&*) z@p#_`RZOnD>QqW&OXxP+F5*a+C)llhTRy*Em9N|u`(M?psvrMsG5H+t3`!}uRw{Nn zn)bdOqfDk%t5GN%e!BT1_f_*EC}L81I`|C`TnpnEl6j?axt9z>@Z_^^kB$y4h%jTO z%wvV@v`)i92~tH`ibBs(ilpH;R7&4pE^{;pO1Uy8rl_V6wy{RaW`_UsW5$GuoeWnzlLq~HUxPS@#8$P3MC~2E zt@lz+OSvj>)-W8rsP3|gjB4a0LUynnn~Vv>PJ{u+fyY&q7-i?l3)-eH_5`)*%46=v z4Sf-uw=S4^k+cpcA@t{216hx+IyA2zy5CK?@ zxe~uVyN%~4=5LE03cH!JtLUCTIZ!6Fi|!38eqUbO&K=v%Wi;0TY8Ok$Z*$Y}_AZpy z$URG5hwRU1za#Jkn-eMY8!a~fUb*XgzoaJ7j;ezS(P6n%@x?a^|IQ&}1SYp#N5 zY?0l@xl60Ez$l^;j!g&~GpZWyPdSen9L6>33#Hu!UWxY4t2%YQBi?a|<(bjBK>Tzj znvEKdJq2C@hTO+5{dA4O_UkuBT$f}ZiDYs15@gAlXwZ~AkMphTV~4p>F;@+hK`JqR zUr3<_zp>gzoN&dv4Evi%ut@fy%P|q~o|nwxR;NV6fb1CTr__eJK82@+OLRCYLiYa9 zIf(nIOu>pumQY5BB6v%MPs83AU7fa8Za5A55aZF+OXLxNB}?f!2jf`qN&-czY7+!e zBf-7PeXbZAbDA$w_5$N)7{jd5xmx^CxJ-Rdg+MmVwGeX=!7i-Mr!d%^>9=a4 zLT;!7h-CEa$3pLnsohk5dZV>VeFsKo(k@-TzJz^=?{U2lCwB}Ng-nd`mya9gcX{3H z4Gl$NahT8UENj^jFdL8N14Fe_CWE}*3o)ef{{3NGEbpJYw>uq~!Z?TgJMmHnqt}r1 z2@>`QQ&I?kkOxb|tw&Bg1^hk)BtnisEDk?!pu&wu))OzM@0$_dhm@{(fJJNlP!EwQ z)12@mOHMHT+sHDuU{Lctd5UE-un!Ob|C=9nB`?SYHw8!jhSma#raFMf4)x^hydbQe5ubLMv<^Y`fxy->&s`ZsvH4KB9lPODEABb9>lmFl zQ?~QikCU7~LZz8`{Lu;SNnZgAE}{I|d^+4ob=l_j_-cI9!FOR~SWd6O;?;nkrcQ@c zDo@Ii8n!13kIho~`S#$d`}OjlZZD_7jJ=G|{r}VTWC7CMi=|9ZuZ?}qMD65kDhXI8 z35?5zogzp+u@{qwBBm+J+ZF71o;5>V=&g_{3V?Nq8wA&8zrSbRJ6($4o*8x#e}g!f zuAH<78rm+=buT1KFquXgMnHNzNGN#_hxk0Y6_yjd&-|6%P)FI&cN|o*=uCl#pW^gG zq5b0GRW6o9Ap*1uHZ+&8a!HZQHhO zcWT?ycIv5Zewlmkk1s2KqLr1LH+w((AWZa9a$`?N6t!}&~h2D1?u%CX1Yk?H?xn+vSRhp3G4z_Pg0 z*fD*HD}bVjxa#{x@(v*LU~{}eUb*Np=oT4g0m70b;4GZvdjcFc*SXLc|J)1lT>VU` zR_-#Cnk+mYVb++f5CVh$n}k1su(zQSm^i>6^dtT+iIR*8=`?*^eMQ)M0SfX_%t=DG zl{~bl@m?u%h0Kg85gBDA-pHv^xh$!uDb`RSs&1xu*v8}SGqtf|0(XdQ4j^~?6iY2T z29+8QSqp3#jn?EixWSxjd_P#sL<}j6Gy>55!-FN5L@emxXP**=brcg%FTr{z4$t8> z$@gEOg-IFu%aETiyT9kn3Dsw8A;O#&NQT}&m|ZW@4-IY$;aj6=*>QK-q{we0cTvwxqbmcMWceh*2la6B5 z^}e2_>3doN)mvlZ1Me$|Vh+auXp`WuY-qxyJjqx^=oGM&V?(pvEwb8)IHtI=4ct>@ z)L8_Ua1U2|wq@Q%onv#KYOSwBksHP!l=BPUF#mvm;Qkpv=_-03M3UZCGZZFPWP2$R z6J4#gUa76Yhu|u_A#oj}2fZBR*<_p1&K#Rr7~&#EDXBMxP#(bRu-wI%BNIptVdk?H zZ6XO*4&GN9$$*e2QLtfDQj5@HpQCB|IeZyPE7d+|wG27pZ)+%?9u3}=svi^g z;;08=R0y4iOlm{&Y=fnNsptGk0wFr@$Vw|rw~M-XbXVrv3+FZK(RB|G8Wo%j!Yr%u z(;qR^Ml%xDC0%}dzFZd|O|g{MeGEeaz7H37oXDo1{!?0ZsN6ejFjndqA$oJ5DCuu) zDU~Ce)yGg-lZMTxh%A9-iH!)3M5AJo++|Zx z_HzhATee9r@hL~Rj}*NXbdkrTN6{RzJA}N_ooBQ85=Jrt0Vl9l-eE`#mb1{JpAifj z6jK`?4ZdROhv8uSw~UT8TMFMKB-d}#kr0gatz*Jnf;r6O*|P1WP~;*O@~o%3^FySi ztNwEL%u-Gxrx^zE;pBMAt6b=Tqv?uQOLC8-fx@=)Q-CqX03`#M*83)AVxyP=Bh-3L zvHLyWKk~ZHon7?m=&)uPCo2S2FiXsO@*)=XFf0J4A`+DTcz<)wbqkBItv%+gY|G~% z(s*?8`=okn8)~lSqwqMo6$|%8f+vB+ryB9R!@rk&5_#@TPeN(kBJ1_Qi%c5hzm>QP z#p4qxbbYU%cBtIJ(-8?ar1^=-o24 zoQQqGRw3U+DTl&OjyVo4h;n?hn0DV9*Pw!Bh?34QKz@!4*M~PvKaK6#MqCrXGQjMK7Ck(BUBADBZqhdC8`{%W*z$> zbUgO<$fgyDXs9$Yu+ccbqdAK86ssU67AkNlH9H7ui(Zu855o@*#oLJ}>$PqeWX!v9y>!A*kvOE-sOUtm1{iFx z1ynZd2K*DSdOp;%sNRk*V=s7ll&LgWRPf>g|^ciT=0$fom2a5(( zs$F#_z;L;0w=Rh?iF%y)pY9LzI<4nGJL~#qa;Li-u z@n&G9rX>jR%AX5xb|FQ@;{1McTXp@o#3cs#N|MNDN6O%*l6TBbm2E^2eZQckBE(?T z0#B=`c-=rU!q9S256uFhb>HZWpPq{7VX z>kxeS0s$h$0jKjgn4Vq8H~)+g-M9fmGBrG~f;B&A1?$g0_@SrMB@YyVp&l=Wx<&eOD2Cb zj>w<{kn;L)X8zx`pS`*e8yUQ==WFNdO_DLVE#5DVPq9L+x?g|u{4PC#!qVXNWM_yB z<`Zha%bdaG_O}612g&XmEckjEH>Up5t-I&e9IdkG9 zk2CIK^{|+#*)p81A97itNOjz&NY^4JTKiBrNY(=HNln@vQGtWrtZG)4#Tuna#})3Z zgbGSV!f9Bgykr&Zr44;PqDHr?+IWmO*tZ#$bvGERn&G~FpvO_L`8ayBtVjQ@{|D}{ zb$2ox4T`8)0Ep!`SmyY$E+OXUbop|@BX@q{utas0@|X&>R;&B zK6M(tPWCQvy+*i_vCYlRQQ&x-R@J%IadUHH0nJg?l4lSeruhwK*FB1C4r!^bj~GaL zr6#cVrX+Lnf7|}}b^&t@rCesaTR<40r^QS*xAOvT5s+Sd-X49<8%8l1b)T0t4Cdpa z_RD_$ymgf}y^AGt(l|iC<98W|F1bI$nPE#T=B6UKQM{aQE#?U)3A;!k4`Z|ETv%Yu zI0TK*$h50olh5-%VR8M8Q!VVf<)B=PcOTk0oLg1h*#p6;KC8=IGroD#f!l6%7)zF! zOC-E(5EBa&l9g%d4G$5uD2>3CxYxmPWX;5I{$9hjSnOCLfn9b+ys?`ieHh^YeCsb|`gdtHvRkkTnh`kpz^C$^G zXYjK^EQTM*f`~v)1rK<9zQE157U&NGytt5U-9X7Eg>j z&koa>>|cilpNAXPbr*|evjO1H_%f~%ZMMH3r&wp%fJFItzK+GH3li}8nq?om{#Pih zChg$`dYeeF>gC5)uAk{~<|co8OaQBHva_5OLY)XYwWQTv8hW|GWM;PcMj;0a;9B{p z7ln=O&tNJim0z*3ub31udl#BbfTZnFMWTUeIG)ZqYjgyn)PO`U5@kBhTZAW3I1Bm^ zNQ7+E%o-L6Mf?!5PFgimGrfJ!4@Bu~KJutx{#4E5NSNKGL}h8z?1=gVB^exs#C+y+ z?<5zN;N^%F80I$GaK}7U8BkSpNV+Q5Zi8vaf*c+ z!K7hiTt?{rx$d~NTrl6^iH9M2Z4E(Sq?C;m3vD4Wc}$aarEqxOG#Gg@BEW{>%doi{g-fNB3A>@%Yx^|IH%s4<%L?)|ZHg`>^iJG=i{ z5xo>CDCk{l>F*}6n*W>N8mAUFdRu2X`qvxb=OE&8m4=(_D^S5S&h4xNg*0lFOFqGo z=r!y6=R+sO^coNS)PMumm>nW+1nmh7w=uZvmAW0bJx|+j*MmsB0v9u0VMe2|1hf24 zx>1-6M^hR5p7;n){}%dXXJ=E;txV}R`i^hZ;I$GuznI-UlHF+>KKS^E#XxJfc=vk|!5(Pf$CD*Ef!A)Z_HrfR$Q z;G>od=ZsdbhpkoKVn)|pdQwJ{&0;g8wK>xk@X=VHrc)QaijDV8~aYGpoK=L&hBwbe$p+iVg;aUA0eQq#XehRYFdOxr*sLL3zXkVkyc*d?*mRqC}I# z7+)twh>h;Po-DmwOpCUIiCnrDJb`5X$ZM_k5yr70v3{6;rE3n#tbN2B*8~|8$MhG#ON*-9igYbHGu&ySvLV5jpq5+9-E~>-*43iNa#a z<@0d7Kb}tkK1HK$d$Yr7Hk&&Med2js%PAO{E_X+oT~8_jZA985j;%1vg|Pzc!n zwrAP*zL>o|UZe{7eZKsS)pQ7IkZ zZ&T~~`jwi)^WgnbL<65z`{!{qAx~iOss@<Ap|)QLB0 zn@S@3xuQ_Da$L45B#BLqf=e$bi) zhNK1^JUq})cg`1NU*n6L98QU6Y^YKT;fVg$wqu2~A2tM+)PPL&Nb#+s5n&X%mO}Vc z*QnfoqG$w;k%xEU7C9UxGfuQYX&}@Z@k~lX(H(JPK}3w}Z{5V%?RgdFrh1Rn#fUv( z-p`#Kl2PVf-hfihzrdl3ArkZzX?;H)KKw(YQi_g=*{LQ~9Yx1T(m4;L`ZO99v$1#* zAOd|AwC3fYN^UwiIqB)^Ki|>7i)3)vX8wR_+j9M_=eEV}e!Y{yYOd_3%%c(&)M+ta zIGx3rR{z(-9O%pBbC=N$5S`YmwU_5tD>Yefs{CG$Hj-&*5T3x|W9oZ%ygQoC^KP`R z!+LAExS!f`$zr!#G@Uwtj1rjiyV!C-YCcx1HLd}$Eqn~CnFcBK7*q(KhZmkV?v)eJ*uUD*DCZeyKgoR zvtre`HNW+a#PMyOtmpdt@je*GvRblVZgr|tFt1tb-a4BnP?XklEZ2&+gc}Q%YhX(# z3(8~;kpCjXSdgI6x~)pnurn#|hJi;CGBb@tJGA6L;Ri&MK=PXvX*gKDN|G7T2^(3G z`iRPn{V0b6rzAZ{fNj7I`!@EA>RAyqK4|Y-Q&Ko0xD{K@)36q-gtR7^VL|4}Qa^XT zCj_{xN((>HLZp~#qFYGEjQAkGi^3*KRNcH`cl0!7I+Unusq1u%2L1ZD zn^BGc^7@y;Ad;6u(cq+%uthSXaQr9SDe<=9uhpp=U=CaMte;lnRJU39&k^+d zob~&3y47oMS4My20ty1F#T@Zd>e~Q3)BVNoHegiwXgnOH<@daI-;1Eqg>$Wn*TmKM zM!<8>xZAR-uhZk{lE&y44^5b4<`#Os-SfM~1*jBy%Dz0wWxkZdpsw}*&WG!CIg`)f zp$KDbSIq8mKDiwui{txxzr*9Sdl#%V`?poOHyUEY=h|wsy@A(jQ<~LksT}y4PBdyk zLBF$=n$&;IFuqe!jW(RBlqnE#c{MAxvyUHj_P}ML0 zDgWs>H$ElH?c0Z;a+749@9Xd{pv84W0E3<2*Xdkn4R6owDALa>Xh;3A6##sCsbVaf zm&|4nw`OBbF@GNup08MXdq|LM@H<09wj-aZ-4+Ux=q3#(x6rmwAspyiC5){^nE!!| zggEma6fztu2*)-z_yF+_^u{t9{y&b8I1gd(!hu{RoFx4CvbRz5aMJ))Yhof|hV7QnF$#GgytNLMnDS zHD~p;_yM0k5enbEzf;q;N9VOA`rdW0DbL&%M1q9DR|ZQ`GV$z?w^YwD?-!>S&pyxGKIw{O3ByeI@;I3*mag!s=lVFNCTa~e`QIBQ>U2V&*VBsILy1REow%E5{ z)b@VpZ}<82dD$!U!eeQK|EI%3tPm2;_k-)*L7+!D*$)^1I^r493W!@TU4VD|7>&*0 z^wWPeE7lQ9qa%=`*XwNFkAj4R{NteYUl4soDIlBM>U5!_r+2=BC*X1ZiSKzZmQbZp_wQUc;d}zx z(N=;O*CEgGNcR0wdiU25;U|ITpMO>*+xZ$8{l4aqsCdmbFNRmuZMP3^CAVkx#q#mm zJZ?X3n(*^nb7DU|!M5%lI44^KCx5?fOD?|4;l{aSvAJQG@h?8rbjlH4tUpb&Bw@*> zH!Sr$65iVdVG`te8pV?>#tMbKGjsr04YEzqa-@7oBwstE$_oUMExzzAJfjKtp|uYv zQYHk@%Jcz+iG4T`KZ$FM!+(Pjo9IFzfM&gEO^vA}QvN}#QzkW548W$Dgl%HOUCI!X zL34>LhaAga+4%lScEzkjx2AKhG=SVgiCsyA@g$51TdzjJ$A#E!{Nqp<)lgI7n_RMg ze2~PA63Yl#V#PtuoF^2iXCYjvPsv>{&n~^mP=cbVCqT1~6f1ruBtL8)1Q7xw)*7RB z941EOR%WW9n+J8MFiEl-a}giZ%dFBM95xH>(6AhNKmB-k%-pOK%M2a+_N24v3M8T( z$E`fC?esxc?YcUHcRAHhaNWmvJ91Mp*L(Nruxd+x7wPij-R8MM>o$g~#$N~tpX)k? zbDIn3HNdRx`}yv{`JvmY`?Kx5UpFb%>6`d(hQ*X2 z>HzcGMu00(Uv>wQ;CsD946g0(Rsum^Z(yYf=l`4ZBH(px=ZrPa^(dV#BEW6A)NGMY zie$%q#NqSSv+CQ=x$@@f=tZ`8g<3>A{`9k1cpOhaGdsj2l*u4C?yZaYdBN`Cwf$z{ zndGX+`*_T0Z(3WF8!I%c*V`f5*4w7m^OgI`KxDNTa{3`hRQ0ek7^}?mO8*BK!I#Mm z$MrXapi-tuOiT&c*ONna71jc_d5t>xMq*8)p-hp8bt1)*b#$fzAc{?Ne zFCvkGOJbg(gg{R@M6@9#QK(Bk5l;let`BIThDI&73ddLCV}t*0G^-PT#V$Ans8`OU zi}x1wY5U&kM}MnCo%nGO9)NgNoX}hqO{*#s0=ia-ib)6w!Ktj%D;~@~UDR%iLb?=s zprim3*pG=Rg;EMo8VIA{fxY`>@L|373&W*GWfk> z-;Rw}hbiPnYr5R7^gEqTLJ{$JGT%f$eA+1H{{54*P$Khhd;78mR`2I)l}*poGDcf8 z)+QC^s4K8SykBnrUxh5fHnZjcN_4wiFE)BSA5MXwT%}&KY5x7`u?Gk|8U{WOpKn)w z?Q&V>xmR2JCBKmPH%qj0D1D&%{=zg@4N55Zl&GWA@hJ4{?r5a~2iesz3`;(rjZ z*pOxC$Zhg&+FxA0w0?b7=`rv)3)1*J zKJzCj3IJ6=s=tfQ<}&@vM1+JMc$%62KuTUq*H$uPx0LNmPhA?b%^23=@P<6c&tU#h z(HFA6XaUV*00Sy36trT9GM>d+6&)%^kGZH6^vhKO?ZFZR#m$I3bn8`#1mZf}ZRnSE zh>iFX=!Wd@(_3w65{Y9fubis1&UY{o3WkzJf3X~Cb=wVm#tST+_b@+%xL@&-kYYM7 zlOJnvd?DJ_^N&=rS*C?hE_=Fxj;$27#B5YSevpQ-`p3W46VvGs!N)-+|KL%=Asnqk zwSJx8B?m|Bi8vqS52VIUaEmTdqkg zV+mw`4)OeukN?eG04?aAv*EtQu~oNskM#Yo;MaP|Fwj_Haqbd|REX!ucUJ7i6DbgJ z+1igXt#f@I*9Qg$jPsUm_~Epuz&^IGpYYk;s6hD915*Wb7|marbk1dS`kI!zdtUD<9P|Z=t`@dzWN#& z`)3(%8{rQxpO~>mKT0_))>+PluI|0YyJDHN+8$ms!hjg#J60aS{V5qdnpXS!b_8^d2!eZ%DUXelKEcYj57h6 zDLX<%e4f^Na))UcAPD34#3Z!XO+!&>;1M)K0 z*dcm!(S()V75H-Bn0(givEtQZ@+e2Y$j24S+4xNoC|qJx#OOg!nKB|*m z>@?Bar{rmtq_WfpFcU;T9>LDBn6&QoKxZ)FC6X1T7z)a0jphkaT3M1sMFvokR>8dc z!|N8oVr_+}`&G*zvdfNL#6IxLjp<;)kv572UzziZB)&N$^SAIGOeetBjo~5JOLp|STw@4$IdX*jUJlcU ze4lq~fW`B6(H4ipIzn+wbnu_fIy@e2c~Cu$Y5 zdb4}6VF5GFrLW7*X_Y8{2ZMg8`q81M{1&`MD^Po5$%(04rvlfx*xI zhQBQpK9ii6nq8(uTYbV6QLGK?;G$*HzP)l;a}JEnY_U#1iRa$X;u?bLImlJOq|@mD z%u$*m*`_#Urc^A3iJ)FmGZ?tP2x*8=8~lG6!e4^d7$FTKzcS~Lz^W8$SdIO~G%h5E z!YLrY)R8?DmSCWRB!p4%CiiqmInKanOF&>Q{(}@?FN}9neC@a4E44`#ylKf7d2El6 zK2FrcZ&hz#rD_Fg$JM$soeaZgGw-OmP`w5bA^$fr|f(~qpI|%nW9E4 zf+ED#96ud8zVD`|(Y%-KcUXPT;y{>iJ706SvH!SmN)`Gd4++$z4=9}{lZ?H;+kLaRC=j6E?d|}KyI%Ll4d%0iouPfe zF)xuxvNdlYu{q?r^qUY@@dKX z<~FNn<>2zgW^x;4-;B+#TBj`c<;|O`aQg8pQ(@gw#Dd&0q*_f zUEMbVPL~AMX46#%86s|W$CagM3+~|QkR@=BtnUQ8Zt;n;0V}5zH8bzeT|bWhiW-4$ z(%`J_+^E?67K(&n952@#`@n>9rc_i?Ha8Y)>K_@4ca|8dFrrCDrfdbHM1c?@pbss! zW=4<4{?4EePm7G6@y%=yrXZV#MB-W#2?Ve3M_9ixPSa)I27nU#XSvOMCjxf0F#de& z!#u~&qjK5JZv(~@0!79+LBeSu6YG+eck$k4id_=mRm)B~;KOHx%s;`4s#Dn(#k2)3 z&~-KyTLR8QjXIvwMNlcjMOKg)dUm@AyOJR` zzum0Ssa6B*|OKBrO{9j3WRGzc1L2K6}4`h^9yneOs+YI4%ae5zcjD3*{P5Q-m) zMkVk4dQt|2nbY5G_VV&G!%J0Yn(63E-o?-YV3xT`n3&owgU!;LHTj9@9#qFe^G ziKA5+>Rm8769r0JJX8;qv=9c=o32a%1)i!@0Vr7K=2lj1NIPgOc#X3q^L6 zvLIy_0WO&^bkbNubz?M3*flfkhLKG$EMEXVJgOk$dUe%;4R*{Jv$;(eb8_$&HXnH3 zcFP^2_#aegv!K$m50^jW`5a!C_BaT0e^bvudy7RP zO;XyNahus};a$O;!EJz!1LNNe_)923N66@{E1^zO+b98}W{c@!>tJ$>@-7st@qx-w z9nQl|b^_6CE}Pr(#58XX2YG$@M5<=@*8v)K%uyE8CQXUbttb?t@h*L(!a@ac zw*Le_kNA2-zv3dMP`_b09{R|=(sYE&cKJ7c+~3Dd&{;)Jgz2Td&k_b#PE>H_TGa(y$pv_OxJxHz7^1k+xU9E~)aPG(Z!NgcmmOLm)%#*)O8(_+`z?%H3t2c5Vp%4^gRUK7koZv&P zn=x(R;Veo08@i2C3w9=3KgwoudHs&2GH&)q|2T{i{M+8L4NM2X{D#cHaW$&l2=Piq zgMQ!)riZ?E0es-~>B?_WjuR-uQ$F~OICo$AV`4zbS1qBI$(Ip&Nq}WbkvUPg#|m*v ztpmn{<8h|eiz0f`DWOM$4}QP*Eeca!d4~2VV*K0bgq@J@lf|SbZpGXRm+MlqPRZld zzFE99?B6*~S>BYTvX@{x1?@KW(x7^*34cK69F%fl++UZUaAF@&sU&)WMldIVOpei? z;SRFmj*)8WWft>LP@)zGP-ykb{#8QxA*m@et%OvnQ-)1as;N_{^~TT|zC{W8)}eDD z=uVY9xegW#g~V%mHN+ZJ#nOS2{@GP*)uLT$h_U`D8=P1T8 zslja3#_$A=mvmtk6Od*eVWciI^9*8Q;|gM&quV!Bxgk1hj@vz$A82uG&kzWRmuVmp z`sdsowqGYre?Od#@i7Rwej1)AZhQRA>AQv5r4RU&hJ)HbP zafp#tbn+Qf9C@V^`ESG+#I?Zmz$O$#^80?!!1=MA3b+ypy12dxNT-HsTb7(#Sj1rE z%TGubSly;7jq8l@>qh($BgYG+p4^DWFlQ=Sp_aUb|6ZJT8BqFH2&kNS^NFb4U}`3T zw4!P@aLY2MQ1$V6?X%wt;4r{v&6=6vp7>D?sM=w_@+GEdd=BST+Ce3Z=*Gz^6Ru^( zO(8U0cghz%i~WNLkx*3^?uVAMf74>FbM~}vqX8pH zl;5Rdu&euhAY#bp6%X+rsm18gtk5wXrjL{1vY{%eZ@`w&^qh@g5I0^!^psP2@1bPT zNo&~jmPAMfIm(<-tq_RI6E!KO#Cy+9=M<|IOSAIDB}7}`%~6ceRo+hk3JyvX1qrU9 zhR(lxRm^7%+TCw-p3E-OwSa-u_hr)H-xq$rN;~}7=awqu_aw4(cQBsx^73-Y3Trx5 zJm*G-egF|b`N4=4N66m^jK24Sum6^D+tUqWHUbl{pc>5t>M^RKFD1llG_pkrO)_n{ z@?OIGuZpPeM4Yk$tQisuOJESK0`hIHtOaZQ=3_urE{t#-9^9-fSD1 z7}hI8bu{>lARvM9x-kfGmC<$aZEP_9w%!P8s2^VYf)FSnu~O$J+>Xa4)q6vpllNHc+~c<oOQW97*CF#AevVk|X>J5Kou$q6~5`11Z ztceKABU=C_fV|-MxNct8ag^C=``d1#IU={{wol}3RYnP@TE+iUtqx?1G*t2n>4onv zDWUP&idvee7!44gxdzJ27FLURVv!mnzjb5NZ3bO|%RfuMf(TRuBu{B_@EHFeNo4FZq=q%#7i0Lmo!BAH_|yBg1U6 zmTs{>fu;r?0Ksf&VNwSG#H4t<> zkm6BzC4}gtW@aAE{dW?9RBo1zqN0#Q)BuwvrF`}pkko+Vtl$1~*GpxJs2&}%v(Ij; z^YP+0F!Fm$XR;gkeXtkEMd}yE7f0n5HSLea)oIjizg_iRkzV)M*X9$KT?z4i#Ap!zzVFVxF6}!x z$!+aw(b2JOs!1RA*BaV9TB9{%fktj>))H0~5)wnV5fTmz$WEv}AP0*KID0B{U9RZF z`mJ8r|5>N)w#?r$EAWY=egCr1C==a&50Ltb<`dbh!Q)#ji zHbhlZf{s`~bFv5i(+L;TGbqK^#hr7FpU_X>iI$@(`lQ^nNAk z7!0B&Ma^-ZXR0w7*&05yEbv~#5mM+O6Xmp=73Y9hVZ0j6LiVe_NmUi;<7WdA8(fk6 z%QDrI28RZ_mPz^rk zFm~+W6q~y0zVqTFepQ0z%pkxIOYB6cv{*PEzGP3O4rNu+qRd>~3l;w6G5yL?Z~t+e zL+oPuQy477Epm5v7mGj-@2|r6_qGD^fWZ)mbZ=%pS#>3RCc5 z4`E$x;RvYw`I6DmQH6Bo-8jGZAFH)`b-EqOA1-t3P{--%>GMUBU;oZ@GSBmYLMP6w zPPV(Z+iZ83%NPDPEg&-%4Ta7&yxrr4M97DYhxZogr`D@#A`61JEq>vj>jF7fCZFTz z<|f%y@ZdiW{SDt@tJ4K|*ZIXoR8*Ap(HBPH)+or9<^qclCpfIXgxk)+L2xgK&skEN z=y5S2zo)9Qa$-UXmIo&t8wnjf2_0LnRY+=mu#I=RDn2$Ofp*4?LwjF zoQ@K!dd4qXH54_O!h6`iH)^R9fD|3LAcV0dj!;_g60Th_dbhpHw=nXJ+$Zg>i~M`^ z*we2BWhwDeIMZj1u!~9(Ey7BQtn?+3izz+J@>D+RAAB`Yz^E{#W~K~6ds&{v7S#5= zyox3;EZINBK~FoJkPUL&6qa?mpUH~6D=|FmS|nE8wTgtQ115v~{v*j8kXAmU>66UNDM>2xFV@^8rN3^P};oY^(qfjkK-4DfLL-% ziwUb%5x_cXCEUll1*uaAs%D!_u$NT5={1FCPUI;UuE^fxlUJ4%v`*6`(t4c~)1ezJ zs<$*M^TrdQbSP_a8+M$av(o19DV|lFWujN+mnXnPTnYun9Jf*k%%}12@VH)XzVtk8 z>ogdPJ32Ok-YsM2laZ0h$;t}6Uk`HFZbW0V1nNHugY@`*yaKN|0zSAV5=qj?p{`II zTSO>M;US>P02!G8EaPpk#KtF3~Z0vNuDU6FlnkhN65A*i%sRPF1UmuY@XKK7@R|;M3eI^bm zsg*Tq@EPk}uZ9Vn<50`zDRPesfpTgd4;SMKDRhI5hkI-z$%Py7oV?uRG3J&P0JqQ2 z@Se9Lz0p*yb3T`C8C$;m?4GT}mrW!URZ#aL~u zJ+yhv=;#hqK4?ASn=JN|pDI9ib%y>u5ZjG(q?xAJh{=?hmfs$O z3;ILn& zoK|&-LF4A2mBI5{E^IDqB-OF%rDaJnH!Y}PgU$rWlsE)fO57ezX9;)O{%&n)YAWK* zF(xj@5&Cf+9vO+t?Rel13f}v2ga`5(iioGIq9Pm&G57KIMCA{a4jkUa$w@j2ijYzO z8f9lkhq#y+veV$sL<-$cm?odM$9Z=MzUlGtU**BcTU~Bm?(QaL>OJV81K>*o5j{OU zEiDhZ{9d}cx;sR_1_kq3Nk~AYJoiUp;O*>qYif)_*LLQ9|HdPe(bp$57V3auqNYB6 zzS*a~=??&Jh84!d#3cO~ia=9OoXO$EA|i@YJcG?2M*lVxhdo!kre(caJA7QdrQ$+C z0Pb7_04f=Rn7NX#uRv!d@Pig)2|;(hT_BSRw^>+Pws>^4499|>xvpU|{;o=0)_G*o z>~L90e4yjJ(zEv%Gahx*tS4pAe3=<&q!rh)u`ng5kC?T6QLt( z$Yz8xYfO;z3?V-VX{Fuo7V-doCGGtukI^DDr1*u<)59 zTtclsp%ny5iZAARnKIV8rnG8Rowr)IbP0Aa1DN`xGOif-x+i2rBis5ilty+`W3hA3 zN`8$pV`YVz!P&AZWAXqp3hs)$7^wh;R0_OWHw|=dqgq9*ww7r{qmpY|YkPTGNqGNI znX91HhPp$os-jLdwGvzVqY8}-1QRV?N8uhDi6vZOlZB(DJkOye>n-JlbXkWsnb&S` zEHyWC3Hmfkz8Q+$A(y3+Ye}!DT=SloiaeY)mi#fuA{g7lT`<7N$S8?Y&hx70Ng@XK z@aoDyU%!^(d;{dZ-C-Xj^T$Xus7hc&z$6q%42jvlKl_iSKuH&qK7aM@YypEg5x*yT ze%^3=j1e~{wY)2M5w&s#)=uWKC^oYVFyoTlL*nga(7`{U)k z%)~+*AImE5CmDo0(x{)*&$1JR+k-3b$xPc99k_ z)38&~;-hbLrb_SSh!o0WS-7Fks_F-*;@E%_$85kb4R*`TVUf0gY9WG=MvZDS8%59_ zx(5yUi;C^fL01(%i#pPS=)K5r)RgGg0zd=+jUI{dYR1G<4Bx3Q_|>~|XzU}|P&L<8F!n5pTl6$wdrs^SqOp0!lc&(`PSOdAnedTr&1>0VzmrM&SSko zPfX*_#m$ZL11{ckquG+vew&_wVOR*<6DlXDd!8!{iDhYPf`_ef4n-SR7VYmx~wjqL0`+xk8b%i%PO{bCl?LGFQ3+x2Su zLjr^V1B2aoDBNLxBqS{CPp1pKAzUt$f8Y!QJv{=5=lNO#BO{}>e<&ybA~7{3rN*F- z&^uZe78Z7Ae`#rn+v&*rcAQEoA}WfGi3wV3CW{jZ-Rt30?fL%Uft`g#tI-7ZTuEyP z9~&E{8w#0>7MX9Y&E;&Rf6m_@B$_N7ljGN~U(C-(S@w2+dJ*UTbh%vtt(1X*VR2z$ z2*I)$HBd|y1}`ChEB06H2#^rw zVu0StR;4da(7dg}mAo678viW&mo7l%sOH@-&&Bs^QC@U(md|%)-}RjTLe_XNKvG*{ zg#1HTssQ4!$|Rqlq1a^)<04-wC^a*oaSTU%4S+DAD0!xKUC^DGE>hCJ1XCgh=WDqc z>8wQ1A|{7|ju%pvUoVA@0)Wblf{>$eoOgsfOvwReJ@b^9m59;;)#p@|1LuZ9js$(| z!b~ugG0o``tk7&3TZk6=h{B=BPSr!e!a0H5g^6G!+InB}@v`^~wLPkfQkJ|!)lEKn z@ulu?fo@2pf>QQK+_~;VIfGtpW+@l98DRe02d82gj&^t6xZhT_(@c-ece(J-w~Y;& znH2g+QSLPDPUS)lo7+-M)fy{R#T23qx5$i4jTkX{M;f&^J^quW#nByPpKGXGhShxO<>zA50lHSGIr7~ZINMeL6uv>H za|9u3pe~Sw#N)A;^u;ifQ}XiqfF&W|^GXb<^%Hci;fG{$Y_!nV#qYK{%?2=-Z&bp< z!qhhh<4Jm*&gh;Q*T;F7&Hv*fcr=~z! zU}$J)Dg-WtiqMznzJEvU0!c$Ow77PDC_e@8Kpjh@#DZq3)h*zFHS>sxtMkt>j0)ag z#)7xzM2GgheQms7K^rt*F1gL0UgUC;4qP1}lgB>hxL~8XAs?bo z)Z+9v<2)=5&r$#RN@7@6PL?W6?($S+D#+CYmJ_4ftJ)xb9nnTQ{g#Ls)J@7Ue-4P;L_g@BC_4&QqbUw zVK}8LNs2itMuQxGND;sgG_LT38kPT{+s9bx$SjjyH-!a&(F`widN*l zTC2zFd<-H2iL$U1s}Xm+Bna<60_Qk9GQz0aj*Q^5HxvQ1s&ClL{oC8U#l;8gptwuN zCnpgu&tdDY4`*+W7x2Z53=HC0kdTn!XdoA-r;+KPnLxNXS+2sGbFD065If`)*#k)c}GidGS%d8(* zeo|5r3+*z45ctH2La?u|FU(00l0ZKdyX7?BS9!TGc0QgA50r&^51Q`@^d|ERYETJ%ddu~_sus;u=mev-IvwZ!;V8&Ty_^mjuJaHU}QqGWE(C~lEAbI z+jof4$k3nbma>OMmT<%Zp|r;Ho`EI4#{v!U*dj`uMv|=^H3E)8c{qie_02F8xF~e(UfUYO{I|$9?+aG6m zuikf&{6Lx_S5^8v#i%rl;UrB=$Q@JC((?GdFf|cU+1c44&ld?VUnX z?c~q*iwg7f+}zyg(MGME$m}rx(EFfvwhX08sIubC`e6FJopiOVC*O^E z7?7C59>3O)Gf9KbD%-e-rfI!8s4xn%s2jEDa5eC6~)90)mZ|L!onzu>5pM?6g((sgV4)5)5{wtc2Uh-|i;W=w)J61RaOyN~yZ0 zH)<^vshr9g5VCoWKH`R+H&J3ejjCUnrD!ZwG?^2q6B-QK6UnTVrb_xZ{_rs|#*`)H z2L-`KirK1#ggL4nspD$gsL486j-6xTwdnR%Qla*#m}V1^sAah6fMzinz~YI$+IK6= z1~oKTwU|@!C=kizAQJ_L$oeYU@?gp31b*u%b@h7!M< z%_;a>Dy5&ou-)T+L{(AFv!TXa@^YvFCH<30dv0hb#=;KW z%~|JM!%GPK6Jx7)kBvD8m?Cgpgnve%tS$jE}MJECcjPiyw~zbP;0$B|Qtrrw;X{go(WyS0fY z2)Q^rScY8qUvHb#6z?A%7W{5Z5`SyO4g4wYAAj1^SR}|oqZbwBKc1LWmFw_>neWYX zW-pJ+N$;ADGrYzV_S!@D`-{q1^o&pXY*j}>(d_1}ny zsplM5R%uqSMak^)_>X?BdSf(MVDU&buOu;&vR~`*+1NtD<5x$YqBNIH?CqYX6_{;{ z#gSq?P)jiCa@qIZvd`UjmBYsP5S%+I|PX4kleKIFr)n4X8#X`Hzhv;F7 zkjJ~~He0(_IcwQW(o=8-{D(QJpC|Jw!@iOBeol5R*JxmT^O;t^SaNei?G+E2Rm#yD zyNj391f_sYkPonza9Mkxqcc2zz2|XrY;sd&)(VI0Q9+OP7h5|cDVQLXHJHsp{QQBf zG19-uH8I{RFe+1uMg*1~cSKQ{>Xl01K*mfpWqwUn*+M%LcuYtkc4lT~yy)J*$dsMj z+|ZU;n3)kl$lJI&%r};n@;Y(_9vmH^GA6uz8!9pI{Y^rGQoS-n(t0R+b2tyl>EAaN z7x`Q_G!0?NqE1VFDH7l9pFRyDZ$2Pqc|gwpEVQA)*VXmNt!dlpwjmMZ0yABU#t^@` z<^&Su4wGqJe9!d+vWmR1PoY7Dxw(11B<`<~xc@P>aB(Y&be)mQHnXtFgyP^HgD+)j z>wXQ9o&CmjW;L5Td)G?+`t|lZl?rJddm`pbGx6%FW|)U+R+K|l3bb`AEl(B|0ubdN z=fdx5enRDpS?zvOwW_6QoIw~zdU2;imGUx2%~{zy=rcLts+ILDHkXPKkRQ!9#9J{3+W_wLDgB=)%UrhH|&%X1^ov1He+ADjx8!+G9P zcQikXjdZ1wp68PWbWP^m+MQ#)VH76hx~=WrrAS;*AG*I1J3=E5SE;aF`lO<39L#1J zqru~;e?f#8yRTsaZ%Rx`f{D-0&gQ8lovcgWoFgTTHO=Snv&ppzq|#F`MEGk!XtVj# zr>q*OI5m+G5ieO9+Nd@uC@83e?X6~W-C}4ZscC4SSQDrnDH0y`3EW09T@W2!5t%P0 zO@8S3wDMsXFSf^uSXfwOuu!acMJC8x@$%7*hSgRZUY_Dn3FC}hYI@tICnZtwlBqA? znj!iQFiARAp^-Gy)R4~0)1N$SYy|b>L_~qSq}+z&l$2PheLX#|`k^p5qzxUH67>-= ztGcU`)BOhzP^pV;JAN}dDwGC8KK^BNzNa5}v$KQ>nzr=;m+sx$U(QBxw) zy2G@WKJPj8-#QmIs8Y?tz{9Su`+$hm# zZY)(xv$2q7Pi4}`$)iB6ihPsL?+f(alms%vuBn<;mVaom3u#}o1*;qvd|N6g;`x*% ztm~$sd$%lGi51;47pFpd4yQ`KZsm^2 zvX1fivusfzb;cZt&bAhBp&`+X;oVhMqlb0LTF0N*)q687-i|5wu7)t&ks5uRkE`B^ zsOJz$T5&sgNc;{%&X(zEfyNh00dVX^91o_dP7?}ec9 z+r8}f2?-$~A#@V%FWSRzv7zlyk&}C_^-y@z3Y4E+TzsQtJJN&x ziJ4hOTZ$G2EeT$9UO3g4K3io$vh*461h-S#+4+}0CB4<@+FyLg*vozy0ZmL?4S8lY zE12_7HER2ZMDNIUwHD3<9$kuApO)(gl)usbY4i(!s;D6(>B5#9RsYwzd9{XwH>sGH)n@lUzt;{z8+}A--H;fN;al}Kk!xhfo4sW5 zN(tA`D$XYF{BiGs+xx9rV@j9fw^Q7S-{-BoDwMNXyel}Rztm49)jjkdC*LW&cw$U5 zr{Fway14Zx3sWl-t{jUWpEc^}?%cliJ83Aa->J_$REpW~E0vAGDnB|oF|M@I z3c;h9^F5tK+TBYK@%;`WOs1oU7rKeU>I-3$Bl z>C@@nEX*U2FktrR6>iGWOx?_kA%bU8N5$BfRy&{Z-OyK4)2!eh>q9eEXQ(21d3m-s z*rAM?WHMjA(4?yT+~B@*hw*_w!!6VwKum&W5AMr6NFt!}VPR!u!YO2W?3s1hTSF_u z7)gk5{GH4A;+|8e|L~d`a`U>$@52PIUuz&Uj-3^s*oJreeugAIhJC%g>$Z)oZ;!jZW0jbo$F~ zVNucY(vss|#;I&cLce*LK9?eDd2{;EP=SY96Ul4WpLF3#7}Zq`Ja zpprK$tkRz-WGZZ!Y|MPjV+x;Cl9Egs zeY}3oHOzkXxEW6SeY6-t=e|8=x7PD<;_>6hQ~On8pin5aT$>lO508)A!?Mvf)?7Ku z^*Qx?KWlq;AC470fewHbv_#mGYcW0Tse5@^TvSx_^{d2L|3JQ9wrx2i>J|@A$^Cei z-bA4vqoYECf+JmBT|GTG13PPLuUpC4-(tu5{#sMR#>OT<857{=2gn31R8&-yv01^w z#3UqsPpOXtU#4)p8l@Q8#h%WOFeHW$zyu@O%^N-eFjg4(j*<@Mz74s0##jUA=jWgS z1cZg9ybhTaeW*W+n&VT8LIgNCI3Rp?mUea?{2Vhl+$D-%WcjRD6do2PE796@{%ak4 ze1p?+2WUx~8^urMAn1*TX2R652SV`a6o1MY$yA7u8i3+zY=_LUWhz=Ain+1+A?crr zmX?+*dJ~U<<|KXi0HnfluHH+)vX5HON)r&=qhiC4iHYGV3`}=Mvy@_Dq%N$6vhO!G zHok2rEbJK^By0w24c8CddgFysiQG>A9oOPMX;tF49U>FL_u%WGg{5*~Y3 zb1#3I5jT7>>V(YFns)n=0mYh zES|hDKfld*F%9SL?yl3@t?ct*b#--|f#W;y4HkOw1;~KtnPLp>Y}wNgxsXQC9H9C| z9^yhCvofgb>5ZjIzPvcz9M$PFSlMP=$a+Pje|)%1FMP z-*@iWQ>1^6`1W#>dLucR+OXhwb9jBC?5W@qv0xT=e@^H4kbt-&H6ZroSX%yP7$+8v zj=a$}q5+@zo_2tk1-F|EtxQKYXyyij7^2o6VDZoKj3%K7+Mz{zdV8a2vOM^(u(9F! zYpwT30?+isgkg;%l;GD;@dUj>q0YtUD1}UYhRRWUZ)?&iUG`5GdT|P(oZQ@aI5=(J zQ4p~A)?6lA0m7IR_LpXHk}y{{El-qmr`t@By2{Ft-d?+v?{Pq&TD1~lV`Gz(BcJxS zwlaZt>rxUB@cLf3YC@+3I5}Scyz;#~sVXlI6oJ5^^`8j}2)H;qhet)V2VCfS+eMk7 z9&;kqkx&7h`cHd|Y74=vnmb#9(1C?F`QQ8o?wB`ZyVz`Y`}S?Y0P<}E0|WY0mFJ2| z;6H$@l~LY$tf>hR?(XkLmq&TMkOvsbm!_ts;;jVYV%>ve z`m#;Eef##q2l5e=xzvWa_`!UNKg-wJs9$YQ!`<&__8Ok&iW(NQveFfgl*z#?^ab)) zQ9+J<$zq^w(pP}OJxCO^o^SNk@3dDKwdhL@E2?4pmVx&DOKGW0#|h0;lA54g(z4@q z#&+gK{NG0Yx7-tc>z z9O3BZCh5LIzW;q_$hqeOpWJI44ONWV4vzZ#FQaT|U66#Eh3mr<$ zp1C<%+T`S9QcitdtATXzE#GdbbUuW3=ZufEU@F`GzB&&@cgo!O{HP?eyhHdZ-Q`_+XhUf@*^|^33Ps3rZ{(R!#seb2RG5ePQdGuSLoSfv* zsny}q($XkaR8&cUfhce*(uj6%rcp?GTux;r3H@ai^Mb{L2RParV#EqU1 zLVHt6mfymWb4tAINleD8+!C$GxbZRBlo)4V&ub-skSuy_3S|7K+zyHy5oB>YK zh=&3V>@x8KlWQw0z@GCh`%}|=Pbkt!+o)-2)jjNFS>L~Z-{5^*UQq$#*kCq?J%Bly zDMA|TriB-Fjy8u&j2jZ)yg}{{h+|Un+iyXK81Qm9it+LiutN-weyy&qDjSrV$^sZ@ zX#=hg_Xzwxcnknw`C z$-Y<*uhHF{p=%P!AW|uOOrJkTJCXFgs3c97on}*j9OY|eA)oBYCb@$mkGut6Uy1~5 z)7iyk@MDlk)Ah-;`|ebLy=;yvI%RW|J)HtXa(Q`Kb{d9`hZ;Csh%;d-&nqAh5fxSR z7WF}4etxswD`*|SDj0ze5Dvfy=_ji{67cDysIS6H1bvuUSiqOPTxtv34{P25o@ECa z2v}#lPygafsTNSJV`yELc0Q*lxz@wR6Q-;DIEwTq>HnJbV}z9FKK)-Ju>lR3eL3*z zi@Uo!kgv0=U;P2~h)1GVS_s83`aw|9Lv*#KVDKzVOic7b+O9)PWO{}A(6qk3K4n>U zM?6zpOw5cMN|xKV)l%w;;IqKxPGR$4eh=%&{|96D9Mb?2q|iF@vwMoNVuZz*sM3}Gne zN2T?!rUCu&_5*}p3uHDh=dTI=U&d;N0B)({*Ntv3{^Fu3JNbq2Y zR8gl)h?S51WI5VoK=y*zVJ%o9kYbXh{f!|!8JVuS?i&Lc3p;>uLPA0!D5^^cg!o~u*l&?KUo%7~GrzBt09NmH zF6g@U3+O0LI_B@TNHiZcRn;fu&oI*`ySms}SuH^ofPNDP`aVGhcSJrk1F*(@ePvK2 z-O^^`?l8E!ySux)!{F}j&fxCu?(PikFu2R$4g-V3_T9a)@x6C%Y(;ea>L;yEb=Apy zGJ6IvBZ0pG`Rw6&N=HLN2Tz!3^N_l?zmJ0IoL%b@LI*-HDA9ws9$ax2X^*bV$aVS0 ztr)LYeIjMns{)nu{YC!FZ`Ali z(&{Afrr_G+B>U7d4V(k)mq`Zms^n4Rnq-4pgBv?bD63Qhq6=0_j#Nwhncxr9KILPR zXvj9n%3B5Kk#lSc275IvtOv3Xurd!Pugbyf;R}~hSxLzgRdwY?{)}aq3RYVbI4YiC?%6RqSa=bqvxhhl5j#^jp>i zg3d5%rlvr0QN0WfqnbHrsJNKSEhG({H!C@0){LQ26gO)CrJ$vs{z;*1V#*i#2X?Zf zE;2d7`11@jl~>8{NB0$ z#ae~j*eF!n!>ieangrLgV-tQJ%PJ=QwilgNZfCrwxgC~cuE@^F@T``l_u}G(QW0ZB z|K^+&TJiPzEhblywC?yL=T_uqnKnt&i3m1f&oM}eXzJs1p|~p?Fb>KF94th})d7Jh z9yZiJ=wQ)Q%{0gHs*=gftoeNf8{cmzaG=&+5H7RLk(+Lr(R2b$xH$_p-PsmK%m!hQ z)9UtH(M@JnQAvJX!zGkoR(d+r)o-(6$)UlpiFeQ);{8T1t_rbqVP`h)VQ+_Y@Ec1u zHZWIKJWv|%Y{Zc&MDwv|0d^LV;E-q~_#f|AU^p>hVGv&5Hc;n?#kUX=d?&H*ifGuF zSDAZtq%(7NKTdMU3ecI;*{!Jwow2$qp#~~r{RI?R6*}En#iT9Z@pD6YDNEnSN9ntp zg)H&2Xq+wV+}{m^%}?|bc6&taolb=hL6ZzdB99%NoDrpn4yS0sBf_&nQq2gBkz4co z6~s-dAy~Z7Q|k6ZG;kuZhNmAQ>L98XDj(8;ToDzRMcI@j+3#q>Ha0~DdzrWd+II2j zt~*_{b!yfhdmzx{z4*#%<-Ezl3y1I^ZA}xqA;fX~H){S8yF>SFz87Bp&+k%EpUh&1 z5Fo?d3@oT)Zl0oXjwqcYopkvz%4af53)6l!p}lH9k?-4A1X{6aXLbsmF@K( zh+{;-94<3|;pU`B3in2Jp*+8%&q#2R;Cgmn`TlmJ%oS9ayS~&+TxAMP$O?hiLfWd1 zc4t9bjRj!^1fd^?VOk41=+z3Hko&;s#o1`o3X}`oTdrAW73ja{2blamsz#Xg93%Vuv@3t z-O6~HI*F$fr6Jgh$gqvCGS}kn~PnBp%-d>?88}#Eu`6W5}_FRF)oQ z1~rw-?I6>-&K#?SoBr*Wv`U;7K^3yhIRO~gL_wHXY zM&o9z9WkL&ecL=z$Y4L?f*juuV#(Fb6$6!yhoqiHVQ1S&In?yE4d<%tSA8{p2l8=} zRr~ZocJ`f=78n^D7uLiGBUKKPB`ESTKgbOH#7;0JlLJNJp(66Q*iM1E8+<|OeqxGT zCnfCy`D<}cL-Z&UCt-tT3E0VvRd+2i1PfD5ZdP=pG}M#3s#B78UqqL3}FNzvCv@ESPni$~prM!af*V$;O{+pwo3G-O3+w zPPwWu>0`gEqg!8JKebc4z3P;l-W~T5&*gnI%&98iP+@FroGYz8*kB=kzC=$?pWBeb zK>Gl<@7hcB{-Gq0m;3a!)5{CBs3=l=ESY%LG8#6|>m;ViFe2jV@G$bn52xxW04HRa z$VJQHS2H*lsffNj0!&oC;O6U#v_4dd-1U0}2a4&l*QW^rl9w~=-!D(e-KU#41qGw#JukGFVIZVy$bkp&U->Y34YKtH+Fs@~RN98joqu{aOpC27Ei2X< zR$f-KvU+*3U^l80@8O%VVB zf4}@xd?jEiyuQYdrM+O0kKaD#!WS)VP_60hvHqI16ha(6x;K52?xoDXH$WPZB_}6$ z?JbDSiw8}A<^|U05DfceJXap6{p6q{l-gcFJ7*`f?O;SAL{4pjq72iNCv~qv_Azn& z@U!zBSX`d$XCy-ij2$M6+1lINcl-Qwf8CsmZnT44#n7$aJ>A_k6c(){r=8f2j0*Ji zIXgRVg6T=&Ik@5{afk{FbFXK#6l)>U`oPR$~D>RDM?;m7`926HuXaX~|N z!ii@evan?0lOHB-t&*u&t4V=6QDgnwM@m6xy4KO}ilrUcEI=K|dU$vML_}y=S)pVt zo7oENqYr<&JG;7K?k$B`W{0Gv%@)L+Mq@J0u2l+;As6*N32bsM{yN~@;OutjUDq)X z7_Rq--L~QDVrD~FCt`*8bA5||8kVbRXowVN$_4#658!ku^DlFZSWI*T`<cn<>o+2)i@xX7P3qrN=Zhn&78zSBr z>rzPJ#+V2mTs}oI*ORgZGG(6$gp>z`54i`GBYzXwX$e)aF()r5OIVn&%T45l>qA1m zJ(9$&M#)5KNcj7o)N!_3k!@AJOm7M?Ks6uvwJ<-wD8n#`uWYtUde~OL<87|4tXB8N zw?}F(W6m}m9`!|`3F~8E>?0rQ=WjPaah&5jQNutZ*Tdn62`OWwY){+;P+Czuc<(Wt zCjiE_sveT08KJv5yM~PSU9Q6jIHJ@^Ho6Z@35_-aVd~EfKuz6J(*x9XeFT*Mx_Wq| zXGBu5vZBQ+oYyS__>r-YkdQ>vHF6+3aQp(v2z)jd!!CdR%xxh3*rc)30xx+$1<33= z@!_a!DxLL`C~%+mmR<%;h2X`ty69vd@S*PNYsO=sHjv`i$6!)uJ;Lo3@eDC zOQB1qQP9#(!5PlunxqB@iEQ?8cK)@z{CIa)sNCLREHp55 zX>lPzNx>)V8|;~pF|DYfWlaYA9oEngPZef8#&x9myNR*Y#c9SJr~ybxs3LhxGWb|& zjqhhj3j$gzj#yV?(OFA&wVUr5EvN!@JijOc0)i^7g~cb%`9^f{43Q}Ee&tAx=0LVv zOKY~Eu}~B4Raw|XcKH*gZmW5b69V)Q`C$r;!1j++%8l<_2TMvqg=IH1%EnUlD37qFipTMrJ zF9QdAeV(p6L%{qcZ^;r!35vTkan9a;Q+z77e}5{f5cBr-A=l&doY@q~cb9;A5Vo+k zvFZ1HX<{^5_#?~!UN1i2T2+gHG2~!#xVW4;H$1Dh9DlRh|Maj|5QHhR83ha1@AnoS z7Q3BKL0>DfB=;KvSplzPP;DZ8cTd4&m_^$-?g?MBZ;W^jO>A3EU*OGwp9`fEvf^w} zC}22WTNn!9p!%b`=`Ic)3CY9Fcqi3W&vI5-UMVGAa$I(&7jBH0a9ri5hbtRLnb`x+ zI`i#a=LOzvPw8<*B`Iq%N|UDb3Tzok$*HZ437F_y$4}69PN4-NQc|$&7~Iskl2Mc_ zd^;)%3Ui)I4?mFHr6LADZgngACmc#698;6~hej3cpMKMlaaeyl4RQ*5latZyJhs~; z$oT#e#nP9LRRs8RD1W+K!-3BaAB=&sY&wW~18K4rK_ql-;SUZDzRqI!mISJiM75x3 zNPFd8T9Z!)mLA@|=Ev6}Pz*N)X)5rhLgv@Sfkzj6d&0@lOWg*tf&m9e`uTFaEov}L zDw)6Le;ugb@?QywOCmX2YwK{iZW8F~8#;td&rHvb7Lv8LP8jkT6M+&^oUEzerp*l4=TFH#3^6ZlSPA}4L}}>AJ~{GYIuA5Tv2ZAwcg&MWerk-x1!V|RsVFN;8?Xv2 zn)Hyjk`a}URR!!07NfU-YB)>G*uWAKrH-L#sYv2cWIX@iW*Fy#yYpyST*3pW!uj$Q zAgJ;F_M{8ZGd3oUWCuqAK!i=*;VgGTwge{m$4aVOG|GBWd<6#uf&A7ex#Nw;R#r^} z=2bNc7&gE-stx6`va=a$Yfi(3mclo)7@qn-mltwcKd_O`F*1LBlLRA<*Y*K>^?x52 z9UL4SnK8~SDZpFS&jcslcocQMkN!}7;l&)_FD@>|ECVD~*8Fm8M;PZzPI%UND8X?o z=Cza<=Tq#XTYM`)RCCfAx2Re9 z?R6tm+`>(+o?Zhp6J+Qe$!CnPgBC8XnW$NEZdPut{_%dXYh@RIzxL%09dYxg{yUJBed4_FC1+X0 zZ|Y}vb#RZG;gRI@x4js9(WlL{qy+!!s$hcp^90O{*m_6PLII#0=viuMnaRPoBGv0- zsJ*_h{CN88`Ph4upB~p0mnMqpvQBt*j>uLE8nyMS(rILPEZtpC$Ugs6-hj>e^W5e2>p&QM2<&$zJFI8C1x`)qg>vX93OpiqC@e36JojUGJp^(7~x@!{bq zxc?pDlN`Kc4-VGdcIk1ZM{Xtt{N-VwLbO?P8t*ztpBq1JthpiMJ>(_ zu)6UoAbovmzWeg$*v#eKFDxt!sB18ZZHa6lCv=sD((3K&-n3bLnk(Rh%M*nV9U&vP z=0XO#aK0QctMxQ|f1ACM|3HOpFB#(OjJ00e#*>&@%xqqmyr< z2vjc4kf`r)^yFWzt(@+gJQOy~`F+3p>*EehrZLcnB*g>^8y;;t5IO-Ko~%sf7Q`2i zZ#SecOuy}!l0!qmBhdig0H**|4-Y3Rd_g`wa6K@#uu4B?m6%-oKOdRZ$;lZP5_EwY z2|n=fR9s#$7jx$P$Ew46T7i5`ULpF7F1p4}Zuq9&r+DlfQq}Ene zF!=f3Fc9dE>Q;#xNRib*=mJZi$aE&{qCH?X_j7EUj7qgoyg>zDx}KB zZ|NHcNESbi2$BBh-wz*&A5{$r$Pcx9T0}~`-QAaw&8%~_=HJrfsxd4iB{Rvq&38in z7+!=TD~OO>-QUAXOHXjQ$W9^SN;Cya*%3vCiHkF#r_2k4^hamIJa8F|f;bY$s1$_O z82}4Rad4697X3NK?XhAr@7!#RQznd-Z?f*D(40le!V!vv9G_N9U*Fr?* zIymICyl8RXaGGJb;7n|RMn>KR17ad#Sm?L>D3MT3K*=kb6Pq|lFu1Ve;I zcJ7CY+Tbh#O+{xn>`A@E>G%y77dI=L?2_oHBY&s!wOEaMWl2m&C(Gb`rxy>D1Rb7_ zZ)5?&%KX^&gLuBeb{bqSHU^@!w5RdlNl8H$!gGQmw+C7EkNBvVd1a@ez)sH-q!L4eKA~>6Sy#E*PyQ@M?$j>P0<*3< zSn&bQ8R1RF7^hdc_{G_Je6_#6KJCjxxcL<~yo9&UnTu8r$CaJ4HTm=4@Z$Y3@8QPm zY_wWbou-Nm8SL}jDDGW}i+Va$3yrHeZwJef>X6;8--E7{antJ^k>SK4A?458lf&T2 zuBS0dfSXSpg7XJ?aMRt1D2IdD!R|} z_7_Wduh*WLy?1OakJ(1bJ+n`P3u=P?C()AnzkFE+d9&VIct_HN4)IUzZ9n;Zh$bi*azSYieQuul8L#Zc1uuOYH>oo|+oRDFb$)J;HKK9NLn)oBn9WyCmhlfvkH< zBq2R7!djl5B;$ZToc)R*Rcl~e?|)$zhi`l_UY;=tl6RqUUB@XUz_8{tbdHm&pJ_Ru z9fwClFfxUMf`C&Q-0HHA6nXD%@9yBLugav`g&PBq-B=87Wbf8;y@Acae|Ay+1e5w# z-X_;wj3FsK%+}Qkg@p}91QdxTgCI^sw!yThca%TwrbpdQMDE#}=l5mQKSj;CZDrad0B@&> z)=h_o_wJ^8X`{NKFIyR~8*HDS(7WSid#`_(qUx`^k0S$2*kSIIq0NpjwrtZketybJ zzAD>}uMQy#8Pk8pud3XB%-|g(bO3W}pI-tK*;NQEZ!y@;ch!d8eJdlscJkhK9ae2( z16SO_KG$zR4Z7_l0WPtW?>B2@XZyHxa$n1)t#q@2StK&AuBDHQCv%%$+1eaJ7#*wv zfsD1awKo^1v0~#^V_z^r9o~G3%7V=rbj)WThsUV_k6KzlfAlM9)Aju0;yvV z;ar?OzEt$-`E#ydn?9~xAE}jgBJv+y@0ZAPHAPrA@^97hLe^ka=x3={`VUJWwiv`e z&`k?-b1}~@B(sNzersi9rn38=cRh{TUP!|*q&gA{_3Ub%V5EIqD=0*-)O?|_5uTk9 zhmZxDI%#H+YT7h z+$w@ANDKaHV-mQ(TdCqREgtFHR(!Jp&Fc#$(6)sQhM)1oU;4t^8b+t*SRIGMwa_0> z&-@bA(jVEd%s0 zLHi~zZz=!NqN4kg-Dj;Bu8xZsv!*1R`6)NkF0=}R3~@Hf-m3g6rJ||r*0!P=HLChy zw2ig35W42uII!Yf410Uu+6}h38R6}Hv)k*Fo>1T#^D{j?JGuzzyA)9|Y;$_~NeEKU zxFI=nOdJX(>&6R{+)9#8`3F>zkvW*CHwyoSa%Ow9Di4|wiGR-R!h z>L)okxzU`aXNrr>)Km*TKJtPJ5J|%O*3UC4uN)BKE>3%~^`ItGy?BV3O(rdEZBnF- z&CQwV>G6Q`)6){^J~iJti3yuw{P3eGCS9oSaa;OksC~!x({kw5bL!LBIgCBpO!*lQ zsU4%LrNABqql{oDl34*?BFYF~6?OHG#|uqy@i1FKNPe)8^~VwBdkrq6K^5UWi+kjP zy;4(>leZ_RR+zBMGowArSb4F2UYLACgS69hvsZ(>mcG%C^T%j${0-0QCqr!Hb98wp zg#D6oPKZISJasT6q1rmd1lg z#Cn38x6Z!qe{t00sqXFNnu6vqmCqWrK};rk39+yMcCU5_OpeXK{wW@;c!M9z_id?0 zW_f&GNq;}F#~E;O3&ndf!NEKFhEHtD?wRdx1q{xj zEJT@gk1bdA{AVKZvVb(ys;Eg|24*W@cE;ad?ST2D)!-mtWoFAm+U40(htT`LR5_pNiA7NoRXAI<{2$5vi2s~Xa zd3jK7bBzwbg?sYB;4vouPEP+;k?h|WXiBUod&4Bq<#D6?@`teAsdDeEVktplU8573 zuUO|q9kA~1?w;oN!&0{YjmKGb@l06x@HwZR0-(=9!)^yM#5r`f4xCTR7^x_w1xz)= zmQk-b5wmK}WAdHm{$7vW!L@wuvkHUq!%G54y71Q2s z!pRB5P6_LMP|}}G=Qhe>07`G=ZDZq5GPv*u5_|YHXR6uBQRb2lYFtF*U*h_WUT!(d z`_4`QGj&BoeF7z1k+&f%5$tlxh=_e==1)mn9fp*0bSap`8y5D$q9U3PY4ZcH8L>A| zNy+rT+_W>Z_SxBgbS)79yP?t1xW3u3AVy7*Qf_|y?8CD=2C!`Gk7(YiNb!qY*@2WFc~(Xw1XImPdIN9?sl8WgwzB)HHHD zO0zoXdRvZTHM6fAxr{vwB}sd2qWoF5;jd2*4iF`U&xiib;dCBdqgPgE_TA=@gzV@$ ziUQy9av0Hvc)@Sg{6DcBMY(?bB9(V)WAMb5@bAqsD#{}ffPGy>e`&#n>|6y?aOR?U zk}3o<vA1E%iAm#dST{O*77(e){)u6O=VH*Pj z^J|4i;{kj;=!%aLPiW4qrg#RFTvR#97Wd5zx2&cLX}gzy9t zMD;u!&dDTB*?gEPm`~s|Xi7-S7Alp<+jzL^gyXkH44kN&pR2Nsb6(%v{DTwVWn54J z9RZuR+B{|Hdl<0uCkTlS#lcEFRb*{#4QqaXwK%_ed=WtgIwqKLcGg>kOaY`}`+?{A z!?}aoiOO5=(Jm)LeH-ol&pnIxC*=iN=gjNN%XWW2KXxNU2PF}!e>W-EZ@RfI`A2?A zbj`0jlf$EYGHF7=Cf)JxL_k`I77B3FT)%vb>-QVW4!@ zYG(#V~$EG>q3;Btknii6{gD&PH5cknhl?Z5JR^7tJ`%HTxa&wz_Q^h?R_8VhWK?M-@~@l15Rrscu&ntN=k}`cDPwmQUVR(U3OcuK{!AK zRtVQl(|C_DO#Z`o*D#MzfN+nzoNRDW3{Oo5iS#LRGtMdPeoNH|$`rcMP5@vJ?~6{# zALodAi#iR$!ph7HBDMK=l5_JZ0&Jtfp{K8nDgLcqCIa5>)Y~NS#MTgw;f`;gj5Zf7 zQ{(y*V}5O%Hm~(hiMCP~%uJvRvVHGQGKP`v^>mxI>;Y#Z@T^p|f~yar?vhZ+P6k0i zw(tYN5P^95n7L(6@f-qzV5vJ`(oBERj!VI-+x{tf`@u&64E?^Uil#-L z`i3;7t@>io3r1Pl-251XP;N%_FgY?{mYu8KJ>&-iv(+ntHL_M8sgUB+6>}Yi9$iiF zTIYhOmjm&kwx_();_wz_y+N-?%4jSS!d-M!RIs|ig*@=3$b?jF)pW@L7w<4RNv74% zdRqC0E~Hfc2!)~SZGG z3P_2WU3O-H(Vt^bFpxW(Q``2Oy%*nYr!m{{&Gl&*T?-TNb2S5M#X^WjX@#^$gWAMhTOz%mXpNWYP$FCriC zbc~R=kEz3Kp~i}env@tJ()J}`Zi4SvGE+T}yo?Ib-PLW!pm-UnjaH3!oY6CvSx;M9R~ z84Rr{@l;4i0++9zK6-70N*B|bNCaD|d6#K6fvC?$qkl0?5#Xsni|~}Mk4{cF4GaQHbX& zA(&RPwzdXR!j*Grr7TVKd1~ej=AR~f6?q9VpJ9WBulDwYJ{qF@+#WWAIH6UIh~Xa| zQL^&UH8nMXJB4h<66L5Quy+1?-v|?jpDJdz5nNnUf>-nHObHPZeuPla;#5y{b0#oVoE)@*(KG~>Q z9pcLpl@z}nf%yWs(9l52jm(TQv$90Am%p39!y$y_WS~ZHu8NC^0WLEdp&Fss7+%iN z__&R_m!|4Omp+#i0nor=1ID0BsXvi5Uc^%JnigmRvlV&3TYpQ>Ja+B5*27$>EINAB zE<9{H+c7Bv5@FsI-AP$X0b1EHZZ3SX*n zLu=fv+-7Gf%7&STrpKQ&!vLxG>m0j~+UU9BM#D;s50^E9Zq?%EX6u{Zf=A&RJFtF1 zA*}4I&B^jzfBsk){Is{Nb^O=YLU zQ>C@Ur?WeFzqQSYv~e4gH@D|onRwbRBk}WX(b?Op_L44e#p{je~!*j7vMy7AOHIF}eSFwg% zvpaJ4_wv$K$!JgFHuu*S8};t6_iI#7$nyWsDm-_k@Tj(FA$+# zeSIaR`Q(A&pZ72+=}o<=W`2J6K7dP71F~xkvz4DcJu~{ikuGrC8Y5<}+f72ZR{0ber=Nq z`wlWySV(N#U75PA-%5tXLr^bO`TPF7R=2%EGhU#FkEN<%KztJt)MBPv+ipb4bmTV5 zNIHgbaG~YCdQ8#Y-Tg1GhtTt_qlf13DkF6^fPz_|jJ|HOl>OPw*vTZ(Z=vfV(eN3AHYG0`8+R?nQ+|LH;j_|GFz{AOz_pgs zQsxrtMVV`gQ&mw(xx1#Prvs!jiWwx!X(*}u0#8a85;+Yi|B^E~PM^i%S%LghuGp>B zys--Hj75P#n9f%R%Q@D3?I}=^{*k9TTtem^9&7k)&3*hub3Of4`1;9}muqVxk!Bm4 z3vFTwX99!qsud;DAk`)Jqz&}>m5#r`mPgI?SgF{0e9x|-X;(WLUJq0y({Yn)7o9uNa=x=4` zy6W(Vh!fNOM5LXyrG`aDgPw63zuUc_E1WOIv*ybz802VDnw1fQyPD`@Q^zEx|t*z+mk_R-5F)e7HA5S^1vq2JKsAfG%7=!Yicym}O z3kd}6#})e+CZM>GTNg8KO8spCNIZVM?ym7NXy$Revn2JM>lHgl72O!*2=G!kkxR$4 z?L&Hrq3UYJjGI6{w+FHgG9!k1+Dm`VI}}XMs!960lWBR z_3!y!gz17Lk)C_1kUOX(7bAqGLj$yo9 z$}>Ae^E;!?<`)X5N=oW->(coXthM;@#I5X`Ff2e66Dl>i!{u zU%*|J%6*LfO>XeApAXwh_&Ly0|DHCZ!t$r}bwk{jQ+YIp@qvgt#MXt>sgkA~vMi~^ zUxhBIkh{|GE6gW*Le}SvkwLv(WcV zP-F}gY;9fGTW71H@Y==K^JZY1Iq(7h^do8|bma(jYLaOm&hKfJUb%wb$KvD!O{rl{ z$PerGW$og)(CC-(iKfp}o1MKcsAqcl?CdNJu9_y6z2Tnow2n5+)6~(?-d-S!pwcU` zrOxvTMKtCFCkH?Nf!LKuo6Ulb$wQbuJup9OXKm~QT1`WH2g)76ca&TH+aDr6BVj%I zJ*@LkjLaTQ34tTH9UL1zybea(9%F4{TABJomR=}Ow zwk8QT%f9FZ&gQc>?|S65y)zqu0Y3YAyLp$Fmj$_FUENrH#Kgn`oxt_>;bGug;Hwgy zq?Tw0JiR@I3YEmqE&UO5bZC;k+azLY)%=EqE3?aSu(e5MZ=OYfgBw!P)rm`hOer~g zTZoJyLN)AY=j`buKcx>=$Rww!sz~XnNl!~uZd%}3O^Tt#29qH4WQP4#!d;z3%h~g~ zP^^Mybn=^05wUsmqV6S*vImom@zY9LA3z|B>XWfR{pN(FX7YwGr}NAJ&#IcozucR2 zuG`kB#d)|AxZlxc@p|3s^XiaLtc;$MRMX9I7Ot)4+H#p#(W2JH{`RVh(sFS5OC-QV z&GM8}$*miYf9(l`$~_6{q2fYgf08(RW!_xfpZ}ZpJ;HkH@u)HZq4T zHz9$0drv-}V^u!z zDbCGYk@|%fdL-|cJfYcct*#J4gC8JTckvt83%|wIM^{&a5;8>u672%t*8=l|uG5o1 zt*xkb=hvrfysc^Cg@6sH4MYozc<X5_EYA9ZabV*2#)2PB+@d$RfymZV@{3JSAq?WV2cXrxy*$wB^62&`%}!|g$0 zlh6A}FcCu_r-SX*f9D*$4aUctH#TC>UAANOXfX;S#y%(3vF6*+(!K%bIpwa0Tp)>S zwzNAuPLB9+@c^gkd0)ArsHl7%o$cy3Q6mqdUzuC1nDA0N-Qgu?w6l(| z91qhl$Rk3$HI)SHl)-S~H}+9pFOQoHWF;UVg1NkB&W|HsBTyJ&s=Q4*T%W9>NuU6G z5(86$@p>f&a5Yx{u50@kBIh8%vthqWIi7|YAo(GCW4!n&qGaxa5iTGDhJ1;ddNEoZ zTZyPSEoxa1vG{r4#3IR0Sr|bE4@v6BreX&J8}i{ON|M;$yPH}b{}&0-UT`|ZzlYe0 zi!zPD;Y3}MTWvCy+x!i^Jzs(N#DA`u+FSjj%z29>BvF*8XHs)|9I%O0yf3!HEC=R%X+&!0t1$LArnZjn34rbwpGPAm42`K7mS;x#Ns8{@MDrNQ@M z-K0&;))7WGJ(lsv9$eEqRwAM(+F$?T75M#ITg zgYJ9*yIS_rshzhBN&AwjK7pwwn?C4*LqzVSFF1;j&&glsfbT{{ISIe^>xye*d6vs1 zEYD+U1+OnJeeX*e#*dF(Msky{bc7q54M-RlvPgHrBBhyp1#`BR!2hsEARczSDE7>W8w zQ_@G|jdngd(9wZETT6T%W2hLyTtlZaXr|qfsVc7AyQDg$e2-bcmfF734Ut1D=LtFbgnYFiVtWQB&A{5o= z-v%=ni{sJaZ{&)9BUF+F0R;fi0ieJ(TBFuzjs0~Z;Rp}_z#0kwKm+{ep)2ZO?`mf6 zYN+PrXy&5N;Av<3p90!FN^VYoxL-iAaR0z}0L2+Kx%|tX|0>EIX1_%lXp}t;0D$@rQNF_e7G>{d z_ub6d%HHC?IaQLt4I2eC<3SAoK>7#%L-gNx7gsOa|6{>_2DAr&X7vr|hb-8CDsPed zH~1eO{3o8!9t>?AsMHDQz&|g@tNg$5CJuK0x*q>K0RIV3wB1Bf0jjV5f8d~s008hR zC^Z#gwK`~Pd3noKA8WE z|GoDAjlaFCn(=qr{~7^$MvUfqA=UCdlvflKGV#y<>CZ8isJ`55qn z{tpj|4E_!GF#E3h|B+A - + @@ -53,92 +53,7 @@ - - - + @@ -159,50 +74,33 @@

  • About this Course
  • 1 Introduction
    • -
  • 2 A new chapter +
  • 2 From “X” to “Computational X”
  • About the Authors
    • -
  • 3 References
    • +
  • References

  • This content was published with bookdown by:

    The Fred Hutch Data Science Lab

    @@ -384,7 +282,7 @@

    About the Authors

    - + diff --git a/docs/case-study-developing-a-genomic-predictor.html b/docs/case-study-developing-a-genomic-predictor.html new file mode 100644 index 0000000..04ddfb1 --- /dev/null +++ b/docs/case-study-developing-a-genomic-predictor.html @@ -0,0 +1,254 @@ + + + + + + + Chapter 4 Case Study: Developing a Genomic Predictor | Introduction to Reproducible Research + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
    + +
    + +
    + +
    +
    + + +
    +
    + +
    + + + + + + + + + +
    + +
    +
    +

    Chapter 4 Case Study: Developing a Genomic Predictor

    +

    The case study presented here took place largely at Duke University and has at various times been referred to as the “Duke Saga”. At a high level, it involved a research group that claimed to have developed a genomic signature that predicted whether a person would respond to cancer chemotherapy. When others tried to reproduce the predictive results of the genomic signature, they were unable to do so.

    +

    A brief time line is given below:

    +
      +

    • In 2006, Nature Medicine published a paper by Anil Potti and colleagues titled “Genomic signature to guide the use of chemotherapeutics”. The paper claimed to have developed a classifier based on applying microarray technology to cell lines maintained by the National Cancer Institute (NCI). They claimed the classifier could determine which patients would respond to chemotherapy treatment.

      • +

    • Keith Baggerly and Kevin Coombes, two biostatisticians at the MD Anderson Cancer Center were inundated by requests from (justifiably) excited colleagues who wanted to use this technology. Baggerly and Coombes attempted to reproduce the results using the published description but were unable to do so. They were able to reproduce certain results in the paper after deliberately introducing a series of errors into the data analysis.

      • +

    • Since this initial incident, a number of other papers from the same lab were scrutinized and numerous errors in analyses were found, many that one might consider basic data handling and wrangling mistakes. In addition, Baggerly and Coombes found circumstantial evidence of deliberate fraud, such as claiming that certain genes were critical to a classifier even though those genes are not included in the microarray claimed to have been used.

      • +

    • Clinical trials were started at Duke where patients were randomized into different arms of the trial based on the flawed techniques developed by Potti. After numerous scientists wrote a letter to NCI director Harold Varmus, Duke suspended the trials to investigate the situation. An internal Duke panel eventually cleared Potti and colleagues of any wrongdoing and restarted the trials. Time passes and eventually it is discovered by the The Cancer Letter that Potti lied on an application for federal funding about once being a Rhodes Scholar. Eventually, the trials were stopped, but only after much public scrutiny and a series of lawsuits.

      • +
    +

    There is quite a bit more detail to this story, which played out over many years. If you want to hear more about this you can hear about it from Keith Baggerly himself in this nice lecture.

    +

    This saga has been a difficult one to understand from the perspective of drawing generalizable lessons. While it’s fascinating because of the sheer number of problems that occurred, it’s not necessarily clear what intervention could be taken to prevent a similar episode from happening in the future. Problems with reproducibility clearly played a role here given that Baggerly and Coombes were initially unable to reproduce any of the original analyses.

    +
    +

    4.1 Initial Lessons

    +

    While the details of the Duke Saga were at time astonishing, it is difficult to draw any conclusion about what actually went wrong and what approach should be taken to prevent something like this from happening again. Most people on the outside were just speculating about what could have happened and the people who really would know the details weren’t talking very much. Here’s what most people seemed to take away from the publicly available information about the saga:

    +
      +

    • Reproducibility. There was definitely a reproducible research angle to this saga, in that the analyses that were conducted lacked transparency. There was only sketchy code that was published along with paper and data were not immediately available. However, in a strange sense, much of what came to light did so because the work was ultimately partially reproducible. That is in fact how Baggerly and Coombes discovered all the problems. They were able to reproduce the findings after deliberately introducing mistakes in the data. If one went back in time and magically forced everyone in the lab to use R Markdown or Juypter Notebooks, it’s not clear how that would have prevented anything. For starters, everyone within the team had access to the analyses and the data. It’s possible that people outside the team might have discovered problems sooner if the work had been completely reproducible, but Baggerly and Coombes figured things out relatively quickly. Also, that is besides the point: We should not depend on people outside the research team as a primary defense against data analytic failure. It’s not clear that reproducibility is one of the lessons learned from this saga because it’s not clear that it would have made a difference.

      • +

    • Expertise. The basic narrative explaining this saga was that the data analyses were poorly done. Statisticians in particular have focused on the use of proprietary software, non-reproducible workflows (like pointing and clicking in Excel), and incorrect application of otherwise sound statistical methodology (e.g. cross validation). Perhaps if better-trained people had been doing the analyses, none of this would have happened. Perhaps genomic analyses are too complicated for the traditionally trained laboratory scientist. The idea is that this kind of work is “hard to do” and that you need better people (or improve existing people). That is the gist of the summary in this segment from the television show 60 Minutes on the entire saga.

      • +

    • Individual behavior. Anil Potti, the principal investigator of the study, was eventually fired from Duke over this scandal and most would agree with that decision. If Duke had fired him 10 years ago, then perhaps yes, this research would not have happened at Duke, but it might have happened somewhere else, or it might have happened at Duke but with a different principal investigator. So while Dr. Potti was ultimately responsible for the analyses, his firing does not provide a useful “lesson learned”.

      • +
    +
    +
    +

    4.2 New Information Appears

    +

    In January 2015, The Cancer Letter published a blockbuster memo written by Bradford Perez, who in 2008 was a medical student trainee in the Potti lab. He saw what was going on in the lab and recognized its shoddiness. Problems that Baggerly and Coombes had to essentially reverse engineer, Perez saw first hand and immediately recognized them as serious. In fact, in 2008 he wrote a memo to the leadership of his institute describing some of those problems:

    +
    +

    “Fifty-nine cell line samples with mRNA expression data…were split in half to designate sensitive and resistant phenotypes. Then in developing the model, only those samples which fit the model best in cross validation were included. Over half of the original samples were removed…. This was an incredibly biased approach which does little more than give the appearance of a successful cross validation.” [emphasis added]

    +
    +

    He further wrote,

    +
    +

    At this point, I believe that the situation is serious enough that all further analysis should be stopped to evaluate what is known about each predictor and it should be reconsidered which are appropriate to continue using and under what circumstances…. I would argue that at this point nothing…should be taken for granted. All claims of predictor validations should be independently and blindly performed.”

    +
    +

    The memo was ignored by Institute leadership. Nothing was stopped and nothing was changed at the time. Perez eventually took his name off a series of papers and left the lab.

    +
    +
    +

    4.3 Lessons Learned

    +

    The Perez memo is critical because it fundamentally changes the narrative about what went wrong in this entire saga. Yes, genomic analyses are “hard to do” but clearly there was expertise in the lab to recognize that difficulty and to recognize when statistical methods were being incorrectly applied. The problem was not a lack of training, nor was it simply the result of a few honest data management mistakes here and there. The problem was a breakdown in communication and a total lack of trust between investigators and members of the data analytic team. Perez clearly felt uncomfortable raising these issues in the lab and wrote the memo knowing that he had “much to lose”. He thought the problem in the lab was that statistical methods were being misapplied, but the real problem in the lab was that he did not feel comfortable discussing it. A breakdown in the relationship between an analyst and an investigator is a serious data analytic problem.

    +

    It’s possible to imagine an alternate scenario where a data analyst like Perez sees a problem with the way models are being developed or applied, mentions this to the principal investigator and has a detailed discussion, perhaps seeks outside expertise (e.g. from a statistician), and then modifies the procedure to fix the problem. This is a process that happens pretty much every day in many labs around the world. No data analysis is perfect from start to finish. Changes and course corrections are constantly made along the way. problems that can be traced to data collection can be raised with the principal investigator. When results are given to other investigators, sometimes the results don’t seem right to them and they will seek clarification. Mistakes can be fixed and results can be updated.

    +

    When the relationships between an analyst and various members of the investigator team are strong and there is substantial trust between them, honest mistakes are just minor bumps in the road that can be uncovered, discussed, and fixed. When there is a breakdown in those relationships, the exact same mistakes are covered up, denied, and buried. A breakdown in the relationships between analysts and other investigators on the team generally cannot be fixed with a better statistical method, or a reproducible workflow, or open source software. Recognizing that this is the problem is difficult because often there is no easy solution.

    +
    +
    +

    4.4 Summary

    +

    The data analytic lesson learned from the Duke Saga is that data analysts need to be allowed to say “stop”. But also, the ability to do so depends critically on the relationships between the analyst and members of the investigator team. If an analyst feels uncomfortable raising analytic issues with other members, then arguably all analyses done by the team are at risk. No amount of statistical expertise or tooling can fix this fundamental human problem.

    + +
    +
    +
    +
    +
    + All illustrations CC-BY. +
    + All other materials CC-BY unless noted otherwise. +
    +
    +
    + +
    +
    +
    + + +
    +
    + + + + + + + + + + + + + diff --git a/docs/from-x-to-computational-x.html b/docs/from-x-to-computational-x.html new file mode 100644 index 0000000..2200c71 --- /dev/null +++ b/docs/from-x-to-computational-x.html @@ -0,0 +1,254 @@ + + + + + + + Chapter 2 From “X” to “Computational X” | Introduction to Reproducible Research + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
    + +
    + +
    + +
    +
    + + +
    +
    + +
    + + + + + + + + + +
    + +
    +
    +

    Chapter 2 From “X” to “Computational X”

    +

    What is driving this need for a “reproducibility middle ground” between replication and doing nothing? For starters, there are a lot of new technologies on the scene and in many different fields of study including, biology, chemistry and environmental science. These technologies allow us to collect data at a much higher throughput so we end up with these very complex and very high dimensional data sets. These datasets can be collected almost instantaneously compared to even just ten years ago—the technology has allowed us to create huge data sets at essentially the touch of a button. Furthermore, we the computing power to take existing (already huge) databases and merge them into even bigger and bigger databases. Finally, the massive increase in computing power has allowed us to implement more sophisticated and complex analysis routines.

    +

    The analyses themselves, the models that we fit and the algorithms that we run, are much much more complicated than they used to be. Having a basic understanding of these algorithms is difficult, even for a sophisticated person, and it’s almost impossible to describe these algorithms with words alone. Understanding what someone did in a data analysis now requires looking at code and scrutinizing the computer programs that people used.

    +

    The bottom line with all these different trends is that for every field “X”, there is now “Computational X”. There’s computational biology, computational astronomy—whatever it is you want, there is a computational version of it.

    +
    +

    2.1 Example: Air Pollution and Health

    +

    One example of an area were reproducibility is important comes from the area of air pollution and health. Air pollution and health is a big field and it involves a confluence of features that emphasize the need for reproducibility.

    +

    The first feature is that we are estimating very small, but very important, public health effects in the presence of a numerous much stronger signals. You can think about air pollution as something that’s perhaps harmful, but even if it were harmful there are likely many other things that are going to be more harmful that you have to worry about. Pollution is going to be at the very top of the list of things that are going to harm you. In other words, there’s an inherently weak signal there.

    +

    Second, the results of a lot of air pollution research inform substantial policy decisions. Many federal air pollution regulations in the United States are based on scientific research in this area and these regulations can affect a lot of stakeholders in government and industry.

    +

    Finally, we use a lot of complex statistical methods to do these studies and these statistical methods are subsequently subjected to intense scrutiny. The combination of an inherently weak signal, substantial policy impacts, and complex statistical methods almost require that the research that we do be reproducible.

    +
    +
    +

    2.2 The Data Science Pipeline

    +

    The basic issue is when you read a description of a data analysis, such as in an article or a technical report, for the most part, what you get is the report and nothing else. Of course, everyone knows that behind the scenes there’s a lot that went into this report and that’s what I call the data science pipeline.

    +

    The data science pipeline starts with the scientific question, moves to measured data, analytic data, and computational results. Along the way we require processing code and analytic code. Results are then assembled into figures, tables, and numberical summaries.

    +

    In this pipeline, there are two “actors”: the author of the report/article and the reader. On the left side, the author is going from left to right along this pipeline. The reader is going from right to left. If you’re the reader you read the article and you want to know more about what happened: Where is the data? What was used here? The basic idea behind reproducibility is to focus on the analytic data and the computational results. With reproducibility the goal is to allow the author of a report and the reader of that report to “meet in the middle”.

    +
    +
    +

    2.3 Elements of Reproducibility

    +

    What do we need for reproducibility? There’s a variety of ways to talk about this, but one basic definition that we’ve come up with is that there are four things that are required to make results reproducible:

    +
      +

    1. Analytic data. The data that were used for the analysis that was presented should be available for others to access. This is different from the raw data because very often in a data analysis the raw data are not all used for the analysis, but rather some subset is used. It may be interesting to see the raw data but impractical to actually have it. Analytic data is key to examining the data analysis.

      2. +

    2. Analytic code. The analytic code is the code that was applied to the analytic data to produce the key results. This may be preprocessing code, regression modeling code, or really any other code used to produce the results from the analytic data.

      3. +

    3. Documentation. Documentation of that code and the data is very important.

      4. +

    4. Distribution. Finally, there needs to be some standard means of distribution, so all this data in the code is easily accessible.

      5. +
    +
    +
    +

    2.4 Authors and Readers

    +

    It is important to realize that there are multiple players when you talk about reproducibility–there are different types of parties that have different types of interests. There are authors who produce research and they want to make their research reproducible. There are also readers of research and they want to reproduce that work. Everyone needs tools to make their lives easier.

    +

    One current challenge is that authors of research have to undergo considerable effort to make their results available to a wide audience. Publishing data and code today is not necessarily a trivial task. Although there are a number of resources available now, that were not available even five years ago, it’s still a bit of a challenge to get things out on the web (or at least distributed widely). Resources like GitHub and various data repositories have made a big difference, but there is still a ways to go with respect to building up the public reproducibility infrastructure.

    +

    Furthermore, even when data and code are available, readers often have to download the data, download the code, and then they have to piece everything together, usually by hand. It’s not always an easy task to put the data and code together. Also, readers may not have the same computational resources that the original authors did. If the original authors used an enormous computing cluster, for example, to do their analysis, the readers may not have that same enormous computing cluster at their disposal. It may be difficult for readers to reproduce the same results.

    +

    Generally the toolbox for doing reproducible research is small, although it’s definitely growing. In practice, authors often just throw things up on the web. There are journals and supplementary materials, but they are famously disorganized. There are only a few central databases that authors can take advantage of to post their data and make it available. So if you’re working in a field that has a central database that everyone uses, that’s great. If you’re not, then you have to assemble your own resources. We will discuss the tools for doing reproducible research in the third course in this series.

    +
    +
    +

    2.5 Summary

    +
      +

    • The process of conducting and disseminating research can be depicted as a “data science pipeline”

      • +

    • Readers and consumers of data science research are typically not privy to the details of the data science pipeline

      • +

    • One view of reproducibility is that it gives research consumers partial access to the raw pipeline elements.

      • +
    + +
    +
    +
    +
    +
    + All illustrations CC-BY. +
    + All other materials CC-BY unless noted otherwise. +
    +
    +
    + +
    +
    +
    + + +
    +
    + + + + + + + + + + + + + diff --git a/docs/index.html b/docs/index.html index 7f6b992..8ab6b66 100644 --- a/docs/index.html +++ b/docs/index.html @@ -53,92 +53,7 @@ - - - + @@ -159,50 +74,33 @@
  • About this Course
  • 1 Introduction
    • -
  • 2 A new chapter +
  • 2 From “X” to “Computational X”
  • About the Authors
    • -
  • 3 References
    • +
  • References

  • This content was published with bookdown by:

    The Fred Hutch Data Science Lab

    @@ -244,7 +142,7 @@

    Introduction to Reproducible Research

    About this Course

    -

    This course will introduce the learner to the basic concepts of reproducibility and replication of scientific research. The premise of the course is that reproducibility failure can be traced to many different root causes and being able to identify these root causes will provide the basis for preventing future failure. The course will provide case studies from three historical data analyses depicting what constitutes a failure of reproducibility. The motivating case studies come from (1) the Harvard Six Cities Study of air pollution and mortality; (2) the National Morbidity and Mortality Air Pollution Study, and (3) a study describing the development of microarray-based genomic signatures for predicting chemotherapy response. These case studies highlight various aspects of data analysis that can lead to reproducibility failure, such as workflow management, data analytic modeling choices, stakeholder engagement, team organization, and communication.

    +

    This course will introduce the learner to the basic concepts of reproducibility and replication of scientific research. The premise of the course is that reproducibility failure can be traced to many different root causes and being able to identify these root causes will provide the basis for preventing future failure. The course will provide case studies from three historical data analyses depicting what constitutes a failure of reproducibility. The motivating case study comes from a study describing the development of microarray-based genomic signatures for predicting chemotherapy response. This case study highlights various aspects of data analysis that can lead to reproducibility failure, such as workflow management, data analytic modeling choices, stakeholder engagement, team organization, and communication.

    Learning Objectives

    Upon completing this module, a learner will be able to

    @@ -256,14 +154,14 @@

    Learning Objectives

  • Identify organizational or analytic problems that can lead to non-reproducible research

    • -
    -

    0.1 Available course formats

    +
    +

    Available course formats

    This course is available in multiple formats which allows you to take it in the way that best suites your needs. You can take it for certificate which can be for free or fee.

    diff --git a/docs/index.md b/docs/index.md index e166b2f..a0c6efe 100644 --- a/docs/index.md +++ b/docs/index.md @@ -13,9 +13,9 @@ output: toc: true --- -# About this Course {-} +# About this Course {-} -This course will introduce the learner to the basic concepts of reproducibility and replication of scientific research. The premise of the course is that reproducibility failure can be traced to many different root causes and being able to identify these root causes will provide the basis for preventing future failure. The course will provide case studies from three historical data analyses depicting what constitutes a failure of reproducibility. The motivating case studies come from (1) the Harvard Six Cities Study of air pollution and mortality; (2) the National Morbidity and Mortality Air Pollution Study, and (3) a study describing the development of microarray-based genomic signatures for predicting chemotherapy response. These case studies highlight various aspects of data analysis that can lead to reproducibility failure, such as workflow management, data analytic modeling choices, stakeholder engagement, team organization, and communication. +This course will introduce the learner to the basic concepts of reproducibility and replication of scientific research. The premise of the course is that reproducibility failure can be traced to many different root causes and being able to identify these root causes will provide the basis for preventing future failure. The course will provide case studies from three historical data analyses depicting what constitutes a failure of reproducibility. The motivating case study comes from a study describing the development of microarray-based genomic signatures for predicting chemotherapy response. This case study highlights various aspects of data analysis that can lead to reproducibility failure, such as workflow management, data analytic modeling choices, stakeholder engagement, team organization, and communication. ## Learning Objectives {-} @@ -34,11 +34,11 @@ Upon completing this module, a learner will be able to -## Available course formats +## Available course formats {-} This course is available in multiple formats which allows you to take it in the way that best suites your needs. You can take it for certificate which can be for free or fee. - The material for this course can be viewed without login requirement on this [Bookdown website](LINK HERE). This format might be most appropriate for you if you rely on screen-reader technology. - This course can be taken for [free certification through Leanpub](LINK HERE). - This course can be taken on [Coursera for certification here](LINK HERE) (but it is not available for free on Coursera). -- Our courses are open source, you can find the [source material for this course on GitHub](LINK HERE). +- Our courses are open source, you can find the [source material for this course on GitHub](https://github.com/fhdsl/Introduction_to_reproducible_research). diff --git a/docs/introduction.html b/docs/introduction.html index 8b59e8e..73255d6 100644 --- a/docs/introduction.html +++ b/docs/introduction.html @@ -30,7 +30,7 @@ - + @@ -53,92 +53,7 @@ - - - + @@ -159,50 +74,33 @@
  • About this Course
  • 1 Introduction
    • -
  • 2 A new chapter -
  • About the Authors
    • -
  • 3 References
    • +
  • References

  • This content was published with bookdown by:

    The Fred Hutch Data Science Lab

    @@ -240,84 +138,39 @@

    Chapter 1 Introduction

    -
    -

    1.1 Motivation

    +

    This chapter will be about the concept of reproducible research and we will cover some basic concepts and ideas that are related to reproducible research. +Before we get to reproducibility, we need to cover a little background with respect to how science works (even if you’re not a scientist, this is important). The basic idea is that in science, replication is the most important element of verifying and validating findings. So if you claim that X causes Y, or that Vitamin C improves disease, or that something causes a problem, what happens is that other scientists that are independent of you will try to investigate that same question and see if they come up with a similar result. If lots of different people come up with the same result and replicate the original finding, then we tend to think that the original finding was probably true and that this is a real relationship or real finding.

    +

    The ultimate standard in strengthening scientific evidence is replication. The goal is to have independent people to do independent things with different data, different methods, and different laboratories and see if you get the same result. There’s a sense that if a relationship in nature is truly there, then it should be robust to having different people discover it in different ways. Replication is particularly important in areas where findings can have big policy impacts or can influence regulatory types of decisions.

    +
    +

    1.1 What’s Wrong with Replication?

    +

    So what’s wrong with replication? There’s really nothing wrong with it. This is what science has been doing for a long time, through hundreds of years. And there’s nothing wrong with it today. But the problem is that it’s becoming more and more challenging to do replication or to replicate other studies. Part of the reason is because studies are getting bigger and bigger.

    +

    In order to do big studies you need a lot of money and so, well, there’s a lot of money involved! If you want to do ten versions of the same study, you need ten times as much money and there’s not as much money around as there used to be. Sometimes it’s difficult to replicate a study because if the original study took 20 years to do, it’s difficult to wait around another 20 years for replication. Some studies are just plain unique, such as studying the impact of a massive earthquake in a very specific location and time. If you’re looking at a unique situation in time or a unique population, you can’t readily replicate that situation.

    +

    There are a lot of good reasons why you can’t replicate a study. If you can’t replicate a study, is the alternative just to do nothing, just let that study stand by itself? The idea behind a reproducible research is to create a kind of minimum standard or a middle ground where we won’t be replicating a study, but maybe we can do something in between. The basic problem is that you have the gold standard, which is replication, and then you have the worst standard which is doing nothing. What can we do that’s in between the gold standard and doing nothing? That is where reproducibility comes in. That’s how we can kind of bridge the gap between replication and nothing.

    +

    In non-research settings, often full replication isn’t even the point. Often the goal is to preserve something to the point where anybody in an organization can repeat what you did (for example, after you leave the organization). In this case, reproducibility is key to maintaining the history of a project and making sure that every step along the way is clear.

    -
    -

    1.2 Target Audience

    -

    The course is intended for …

    +
    +

    1.2 Reproducibility to the Rescue

    +

    Why do we need this kind of middle ground? We haven’t clearly defined reproducibility yet, but the basic idea is that you need to make the data available for the original study and the computational methods available so that other people can look at your data and run the kind of analysis that you’ve run, and come to the same findings that you found.

    +

    What reproducible research is about is a validation of the data analysis. Because you’re not collecting independent data using independent methods, it’s a little bit more difficult to validate the scientific question itself. But if you can take someone’s data and reproduce their findings, then you can, in some sense, validate the data analysis. This involves having the data and the code because more likely than not, the analysis will have been done on the computer using some sort of programming language, like R. So you can take their code and their data and reproduce the findings that they come up with. Then you can at least have confidence that the analysis was done appropriately and that the correct methods were used.

    +

    Recently, there’s been a lot of discussion of reproducibility in the media and in the scientific literature. The journal Science had a special issue on reproducibility and data replication. Other journals have updated policies on publication to encourage reproducibility. In 2012, a feature on the TV show 60 minutes looked at a major incident at Duke University where many results involving a promising cancer test were found to be not reproducible (more on that later). This led to a number of studies and clinical trials having to be stopped, followed by an investigation which is still ongoing.

    -
    -

    1.3 Curriculum

    -

    The course covers…

    -
    devtools::session_info()
    -
    ## ─ Session info ───────────────────────────────────────────────────────────────
-##  setting  value                       
-##  version  R version 4.0.2 (2020-06-22)
-##  os       Ubuntu 20.04.5 LTS          
-##  system   x86_64, linux-gnu           
-##  ui       X11                         
-##  language (EN)                        
-##  collate  en_US.UTF-8                 
-##  ctype    en_US.UTF-8                 
-##  tz       Etc/UTC                     
-##  date     2023-04-11                  
-## 
-## ─ Packages ───────────────────────────────────────────────────────────────────
-##  package     * version date       lib source                            
-##  assertthat    0.2.1   2019-03-21 [1] RSPM (R 4.0.5)                    
-##  bookdown    * 0.24    2023-03-28 [1] Github (rstudio/bookdown@88bc4ea) 
-##  bslib         0.4.2   2022-12-16 [1] CRAN (R 4.0.2)                    
-##  cachem        1.0.7   2023-02-24 [1] CRAN (R 4.0.2)                    
-##  callr         3.5.0   2020-10-08 [1] RSPM (R 4.0.2)                    
-##  cli           3.6.1   2023-03-23 [1] CRAN (R 4.0.2)                    
-##  crayon        1.3.4   2017-09-16 [1] RSPM (R 4.0.0)                    
-##  desc          1.2.0   2018-05-01 [1] RSPM (R 4.0.3)                    
-##  devtools      2.3.2   2020-09-18 [1] RSPM (R 4.0.3)                    
-##  digest        0.6.25  2020-02-23 [1] RSPM (R 4.0.0)                    
-##  ellipsis      0.3.1   2020-05-15 [1] RSPM (R 4.0.3)                    
-##  evaluate      0.20    2023-01-17 [1] CRAN (R 4.0.2)                    
-##  fansi         0.4.1   2020-01-08 [1] RSPM (R 4.0.0)                    
-##  fastmap       1.1.1   2023-02-24 [1] CRAN (R 4.0.2)                    
-##  fs            1.5.0   2020-07-31 [1] RSPM (R 4.0.3)                    
-##  glue          1.4.2   2020-08-27 [1] RSPM (R 4.0.5)                    
-##  hms           0.5.3   2020-01-08 [1] RSPM (R 4.0.0)                    
-##  htmltools     0.5.5   2023-03-23 [1] CRAN (R 4.0.2)                    
-##  jquerylib     0.1.4   2021-04-26 [1] CRAN (R 4.0.2)                    
-##  jsonlite      1.7.1   2020-09-07 [1] RSPM (R 4.0.2)                    
-##  knitr         1.33    2023-03-28 [1] Github (yihui/knitr@a1052d1)      
-##  lifecycle     1.0.3   2022-10-07 [1] CRAN (R 4.0.2)                    
-##  magrittr      2.0.3   2022-03-30 [1] CRAN (R 4.0.2)                    
-##  memoise       2.0.1   2021-11-26 [1] CRAN (R 4.0.2)                    
-##  ottrpal       1.0.1   2023-03-28 [1] Github (jhudsl/ottrpal@151e412)   
-##  pillar        1.9.0   2023-03-22 [1] CRAN (R 4.0.2)                    
-##  pkgbuild      1.1.0   2020-07-13 [1] RSPM (R 4.0.2)                    
-##  pkgconfig     2.0.3   2019-09-22 [1] RSPM (R 4.0.3)                    
-##  pkgload       1.1.0   2020-05-29 [1] RSPM (R 4.0.3)                    
-##  prettyunits   1.1.1   2020-01-24 [1] RSPM (R 4.0.3)                    
-##  processx      3.4.4   2020-09-03 [1] RSPM (R 4.0.2)                    
-##  ps            1.4.0   2020-10-07 [1] RSPM (R 4.0.2)                    
-##  R6            2.4.1   2019-11-12 [1] RSPM (R 4.0.0)                    
-##  readr         1.4.0   2020-10-05 [1] RSPM (R 4.0.2)                    
-##  remotes       2.2.0   2020-07-21 [1] RSPM (R 4.0.3)                    
-##  rlang         1.1.0   2023-03-14 [1] CRAN (R 4.0.2)                    
-##  rmarkdown     2.10    2023-03-28 [1] Github (rstudio/rmarkdown@02d3c25)
-##  rprojroot     2.0.3   2022-04-02 [1] CRAN (R 4.0.2)                    
-##  sass          0.4.5   2023-01-24 [1] CRAN (R 4.0.2)                    
-##  sessioninfo   1.1.1   2018-11-05 [1] RSPM (R 4.0.3)                    
-##  stringi       1.5.3   2020-09-09 [1] RSPM (R 4.0.3)                    
-##  stringr       1.4.0   2019-02-10 [1] RSPM (R 4.0.3)                    
-##  testthat      3.0.1   2023-03-28 [1] Github (R-lib/testthat@e99155a)   
-##  tibble        3.2.1   2023-03-20 [1] CRAN (R 4.0.2)                    
-##  usethis       1.6.3   2020-09-17 [1] RSPM (R 4.0.2)                    
-##  utf8          1.1.4   2018-05-24 [1] RSPM (R 4.0.3)                    
-##  vctrs         0.6.1   2023-03-22 [1] CRAN (R 4.0.2)                    
-##  withr         2.3.0   2020-09-22 [1] RSPM (R 4.0.2)                    
-##  xfun          0.26    2023-03-28 [1] Github (yihui/xfun@74c2a66)       
-##  yaml          2.2.1   2020-02-01 [1] RSPM (R 4.0.3)                    
-## 
-## [1] /usr/local/lib/R/site-library
-## [2] /usr/local/lib/R/library
    +
    +

    1.3 Reproducibility Guidelines

    +

    Finally, the Institute of Medicine (now the National Academy of Medicine), in response to a lot of events involving reproducibility of scientific studies, issued a report saying that best practices should be done to promote and encourage reproducibility, particularly in what’s called ’omics based research, such as genomics, proteomics, other similar areas involving high-throughput biological measurements. This was a very important report. Of the many recommendations that the IOM made, the key ones were that

    +
      +

    • Data and metadata need to be made available;

      • +

    • Computer code should be fully specified, so that people can examine it to see what was done;

      • +

    • All the steps of the computational analysis, including any preprocessing of data, should be fully described so that people can study it and reproduce it.

      • +
    +

    We will expand on these ideas in the chapters that follow.

    +
    +
    +

    1.4 Summary

    +
      +

    • Replication, whereby scientific questions are examined and verified independently by different scientists, is the gold standard for scientific validity.

      • +

    • Replication can be difficult and often there are no resources to independently replicate a study.

      • +

    • Reproducibility, whereby data and code are re-analyzed by independent scientists to obtain the same results of the original investigator, is a reasonable minimum standard when replication is not possible.

      • +
    @@ -335,7 +188,7 @@

    1.3 Curriculum

    - +
    diff --git a/docs/learning-from-failure.html b/docs/learning-from-failure.html new file mode 100644 index 0000000..ba59204 --- /dev/null +++ b/docs/learning-from-failure.html @@ -0,0 +1,220 @@ + + + + + + + Chapter 3 Learning from Failure | Introduction to Reproducible Research + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
    + +
    + +
    + +
    +
    + + +
    +
    + +
    + + + + + + + + + +
    + +
    +
    +

    Chapter 3 Learning from Failure

    +

    Learning about how data analyses succeeds or fails (but more importantly, fails) is extremely challenging without actually going through the process yourself. I don’t think I ever learned about it except through first hand experience, which took place over the course of years. There are a few reasons for this that I have observed over time:

    +
      +

    • Success in scientific data analysis is usually concerned with whether the claims made based on the results are true or not. If the results feel true, and the analysis appears rigorous, then that’s usually the end of the discussion. Focus is put on the result and what should come next. The underlying idea here is not necessarily misguided: Progress in science depends on independent replication, and any given analysis cannot be assigned too much weight.

      • +

    • When analyses fail, the results are usually vague and confusing. Furthermore, the public rarely finds out about them because they are not published. This is mostly due to human nature: it’s difficult to motivate oneself to write about an experience that was inconclusive and perhaps incoherent. It can also be embarrassing if honest mistakes were made. Publication of negative studies is a separate matter, because a truly negative study is, in fact, conclusive. But often, we don’t even have that much clarity.

      • +

    • In the rare cases where we do find out about data analysis failures, the focus is often on who or what is to blame. In cases where criminal activity has taken place, this is an important aspect. However, identifying who or what is to blame usually doesn’t provide us with generalizable knowledge that we can apply to our own data analyses. The underlying assumption of assigning blame is that this failure was a unique situation that could never have happened if the individual to blame had not been involved. Occasionally, there are cases where there is a clear bug in some software that leads to erroneous results. Fixing the bug in the code will “fix” the results, but even in that situation it’s not always clear that the bug is the ultimate cause of failure (although in this case it is the proximate cause).

      • +
    +

    The case study presented in the next chapter is useful for thinking about what kinds of generalizable knowledge we can obtain from data analysis failures. This case is special because it had serious implications and large parts of it played out in public. While we likely will never know all of the details, we know enough to have a meaningful discussion about the lessons learned.

    + +
    +
    +
    +
    + All illustrations CC-BY. +
    + All other materials CC-BY unless noted otherwise. +
    +
    +
    + +
    +
    +
    + + +
    +
    + + + + + + + + + + + + + diff --git a/docs/reference-keys.txt b/docs/reference-keys.txt index 1d1a661..c0eb704 100644 --- a/docs/reference-keys.txt +++ b/docs/reference-keys.txt @@ -1,28 +1,17 @@ -available-course-formats introduction -motivation -target-audience -curriculum -a-new-chapter -learning-objectives-1 -libraries -topic-of-section -subtopic -code-examples -image-example -video-examples -using-knitr -using-html -file-examples -using-knitr-1 -using-html-1 -website-examples -using-knitr-2 -using-html-2 -citation-examples -stylized-boxes -using-rmarkdown-container-syntax -using-html-3 -dropdown-summaries -print-out-session-info -references +whats-wrong-with-replication +reproducibility-to-the-rescue +reproducibility-guidelines +summary +from-x-to-computational-x +example-air-pollution-and-health +the-data-science-pipeline +elements-of-reproducibility +authors-and-readers +summary-1 +learning-from-failure +case-study-developing-a-genomic-predictor +initial-lessons +new-information-appears +lessons-learned +summary-2 diff --git a/docs/references.html b/docs/references.html index a1b8fdb..e5793fc 100644 --- a/docs/references.html +++ b/docs/references.html @@ -4,11 +4,11 @@ - Chapter 3 References | Introduction to Reproducible Research + References | Introduction to Reproducible Research - + @@ -16,7 +16,7 @@ - + @@ -53,92 +53,7 @@ - - - + @@ -159,50 +74,33 @@
  • About this Course
  • 1 Introduction
    • -
  • 2 A new chapter +
  • 2 From “X” to “Computational X”
  • About the Authors
    • -
  • 3 References
    • +
  • References

  • This content was published with bookdown by:

    The Fred Hutch Data Science Lab

    @@ -238,20 +136,9 @@

    -
    -

    Chapter 3 References

    +
    +

    References

    -
    -
    -Allaire, JJ, Yihui Xie, Jonathan McPherson, Javier Luraschi, Kevin Ushey, Aron Atkins, Hadley Wickham, Joe Cheng, Winston Chang, and Richard Iannone. 2021. Rmarkdown: Dynamic Documents for r. https://github.com/rstudio/rmarkdown. -
    -
    -Xie, Yihui, J. J. Allaire, and Garrett Grolemund. 2018. R Markdown: The Definitive Guide. Boca Raton, Florida: Chapman; Hall/CRC. https://bookdown.org/yihui/rmarkdown. -
    -
    -Xie, Yihui, Christophe Dervieux, and Emily Riederer. 2020. R Markdown Cookbook. Boca Raton, Florida: Chapman; Hall/CRC. https://bookdown.org/yihui/rmarkdown-cookbook. -
    -
    diff --git a/docs/resources/images/02-chapter_of_course_files/figure-html/1YmwKdIy9BeQ3EShgZhvtb3MgR8P6iDX4DfFD65W_gdQ_gcc4fbee202_0_141.png b/docs/resources/images/02-chapter_of_course_files/figure-html/1YmwKdIy9BeQ3EShgZhvtb3MgR8P6iDX4DfFD65W_gdQ_gcc4fbee202_0_141.png deleted file mode 100644 index b865852d5ade5fdde09681f2d60e0c8ef89fd21b..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 17318 zcmZ8pc|4R~)E`Sx8d|BbRnjK>%9edQsK~xlQe+$1 z_ZG`!4I|m#^E}V=e%{afpF4B!_nv#sx#ymH?swchrKg2oOIVA;;qcnWkDbBcINsrK z>_Xh^@JXKfsRKA1$wB+rk+Yr_V?8`2hL?NaO-daS_t+v9dseDCRCa6scr>@uUUBN` z#;y4WcX;SJ#~s1%C=8FfF0t{G&X>1O?jCp6{<6R(d4Ri`hC;7+m2%G#4iV)wYoIPNWYd>5U5T!fu~`=Cjxke8L+NTsZ%9HJ#( z6eJL|j}keEUUd2^JU;B>b{ZSao=k41QZ)~)r{hlPVK8g}vsZ>d$V*P%;K3=2dmxHJ z-jM^yoj0gd2QvAi02>~6`UnbtDH!0nlTrBV#{fPTgKytKW83bG!IuF1dJPO-6X5Uc z#oz@1{+Iv;zY2j5L*b8!BJf%${LXNI--=0pS`*+^F!-vS056Zh%WeSpJsA8=1b!O^ zF9+}+`7rp+0DtK&2G0fX*EKNsn+SXgD*b69z?U?l%6TIO@P9CPFMyBaMB!6H0sc5Q z3V&J(;CFGM@Oyaxeq$gCe>(yoQ!$0?KwxfQFb|QTyD`ZKNcRw|`z`3c3zJ|6m}L8R ztow6hR^NM9882S!&+(wp8tHrY!Kjx+RPGZ&l+`#>OcfhJ)9-d-@@fj|4Z(pwoZf zfiC^<*?w8s5?DJ9am}^(uo9%ZKiXYP0lI%%gLbdF0o^^paM1k?EE=~8{9+4%unkm2 z`G5^|1;J<+u$jP~xJ9gsBYgaB4}oBm3|>vg;Z|Tt)^k$>gB~I;krf7GLii&>RDv-8 z6-I2wp+KhKgE&sHo%qtLE&BD?_m&dvzF7kj46=n?yX8HnlfuXw)%~_+eyIw+O zbw+1Z+J?ZyfitYdC*TkX7E#D>#9r&ypo;iR0O9YfMd5#j0Q@m-6#k$Fz&`=^2l&It zy6<7|qX_(74Bi5PKZe2!iy^u#!r+mev)&eiXGh?R7_v-{K0lj3yz6BNO7qCDb*OEH zuOq`muwjRgVS?x|Eq-L!CRD1FyCBuab*NMu5X1fmLZy0!Ec*2^#9*$eh*SlbR7wbZ zC??eg1ilx8XGh?DQ21D3fIrTT!q*^+-o=T+TO;sMm`Shk^K**@kJnE z|ATMvZbM*T|AXNxkfF%_2Yu0z?#TX!?m|d+*#Ag(Hl({6+Fb`R{5c4{&^-!KZyxH@ zMXn$`MFZ_FwE+Z<^c`ulRKw+e zO*pEc>BDN4A`V||e2UdnMbuDjO#b&RvTv)0#Apy#OdIJ)#HRF53(z6VWs(`)h>@q} zc*zhtjAUJTQN0AOVKtG>xqU8p9eLm?BU-}PD7xW|bxnL^>KkQ5)_NUKF$B)Cr0PNEU9}YxgN#^lm*p##GTafn z_w%T*L{^6;PVm7_Ruh%!#AbBl0Vgu5)ifSdVtEBwRI5>|bHRrt>;xa7xX)3GL?_^N z5xq&@MwM)Rk+cb&W)IDfN$?){5-lrx^nxktLrHY9^DT6Awh!x2_&>53d?bb^&ca*7 z@JQHD3Cw6Gd|6FnqJ%L?+M=X=UDx`31j6?d?yN5`Yo54~s$$WOC4x>O^KJlJnZ$x#6(8l<+ep@wL*Mf==5-$fs%WC2#u9n>i(qVDp!8|YiWhXevqDSlUf zpz854;6zB4!1Vgp#*}2U4*SFj1Fn2Qt>;Ha8Gk}Y?Tt~zCy3D`z{*)>=;GGnbO=fw zL}NOg+x$GL$L(9?WnFu**?(jD`0uV0SsIgY2d1#ds2;iqN6}5HCgTID018FCAi0#w zbrn909@vVi{JqXqlE7w6XgyS*2gh~_7SS>2^_bDlMzzt^*ioPpJ?Jp;2d}7q1Td+; zVp2=qZJQ`x@%Ng0c|PY;N7xU+PE6vSsVLMNDU!iXj7uz(W)y{g-<#wRmNqod$=lN& zQl->r5@z_N(G}eZW>E6LeD<{0Z2i1%;(62KPu`(xeWz3=s1!-bx+ce@?D#pnH2B7O zsqIg}d)Zv^n+}L^aDsDo3>r2UJ7#;GIWhx>DdXllXWng8e5?fzOy66?nRB`*zv~tihvK2>rj$BZ5dxzxdvt=XO~RRF>~=_ z>V!^_GE`cedQBHy9FH1(x(!|UUyX69lpAW$ga}l9L}5jO(rrxZN3mf~?eW5Tn9+}* zZa`UgNZyr1pDOtHoUMih8&9_ON~=`)lrG1FqhlT*F(2ngUVBb znX{_Fmf^e#RmA|h_i&z&&Uw!avp%#&w*_u9arixiFxdO&F;^vCOx<7m3cmp#7MX_n z9B!?VHmQj&e-L#Xod4OcblF_pZd80+(tcYqp>uZ;s@ZRr5Iad-AJNT(+ng3feVXjB z6ImsN`2^t+RZj?Up$A4N3=dn_ zhm#3PZg3NQVVvE|t33Lf*`)ut;!g6+NSmeCK(y{-d588_-|y~8EO7mG#$`2bb0)Hu zO}8nUZ2gVt8X^2M^6li5FtsJ7fy>4pgRqJwOM}KtQw8YDnT+(kf0Od(=hAGSx-Une zJulbQ$sd(k!fp>-%=J|!Zz~FARo~xJtfLdU_27s2+Xaeq{r7t;nKoOm8l#`3UI8)oytX5!pYtlv9eIU()9NFJ2tMDmgT+i8{*6S7KTUW z7y4Eld4ueFOT6#|S^9!ls$#A7oyQ^4W$jCHOLb2Em+SB3=tvbF>&w+u=xG*m_gb9% zZ0&x?l=L)MMy21o~LHxmMYHgMQrORtNF9r$MUW^s6 z@O(ehWaRYz=BmMUG@|`PW1J~ zpUvHq==q@_<_@Rq`lmJ0TR(^(lN!W@dm?l}r;^Fhj+{C#jmV^7VUO;{0)BP^|NcWXr}8ED(s>J? zF>Bu9lDWah-cKWwj<~EowS5oGmaOpHrjSG?52>>&m2s2RzjBzk{`y)OZ_GK{mjDF_ zWPA`_@TqYh+z{$6#y&SlAC9~EbK$fKq^`MxoU-__MJ?iI0k$!%PZa(J^_3FEme}m_ zkk37@qWFKi!NnZr67_whIC|*vr**0>Kf@H{gKm5%9;1jY^Ccv2=u*1b{Cd5H3iV}`J%3+talf%!_fcwoYmQAr@pzHWN|PnA zAwk*L{cod|Xz9xG!bHJO-ooFD(+vyDobR_t7}fUBUcC4s^hdL7tpAEqv1^~-Y;~Zq zrC-bJp5rw?vwPD*KWBcq%)Gs3Q@;t7PqfUjN6DpNeM?Krd{MvA!-{=80#ZWi{v~ET z^kmY&%2!{RE47hZ>bxrNb^842u$g%qDxx}H!KZGPb0ti%L2zy<*~xgkZ``get3^ga zrnFAaC`r*}ut{HC(nrp!YW>{G!-JRCDy2{Ts0^7gl$55=w~)Zx+4QlWW`LR5tg??7d8N zF?hA9)79n0xn*{fvS*R{dDB%kp@QQzM;3Z3=7lDv5odOcGVY zq#d~hYe=$mf0EVX(YN%-F6&>Iy%A$+v^pg-NXTH+OPl7gyUW5Fm_Q`t+9E!*CAV!ttzV(R? z?GHa>{O3Y-l=vX^Ouy;T1|+zFZ?HXLBI(q28*fJ0NAKi%asr{u{ zICt8n$4Pj~`~JOFnozzp=JY0&PKnWa!p5=MD&%44F7J}0&z+V;o%=HreU?+>v37M? zr&BZo7JrJCmTyluNYe`4`g$P7v?y;r=L~0V^NZA+&!&3<$0qKd==GT><={R1uG@@X zc`Uc7>YiAgu9dMBu^~=os7frNxyFS|r`=er>Mskvz%j4BG8d89RP5p8(yk$bpQ$zX z=xGsJy*5^o7F~&ld7?AnU_wVL_tbV^>w6+zX$Vu*!)3uF0%gNG(J(3_s>c3H=o`xhQ^dZgeqKI!ryj1tqQ zJ+3}4$j(PD>{D_dWfZN}uUA{Jp9nIok}UeuX(uRePj(bf$}%sOGR>>)Hrrjv$t<}3 zQ@+R5pp+g0F69Hq&EQi7Zh6`sz4&KCRXvWp#Z^wasxlXoY{H$ZH_*SdA{pwJFOW_^ zS>unGDy_1=`u@ykM-OnD&#&{wFF#N6`*YqVd6wco|0{jwWYAg9mF1;0r*5t2s-GWH zweM#Y4ifgAowf2}|0e!xEuMV$+f&`g%g%DoON+i!1$+HwdtehEcG-1x1x}@2_b!Js zlH+lL4!MoD_|(pM_-xzCCaF;OHY`bR$*03QR&}aMY?DE-%FPelWUqLKo}#xFUsJU= zY0GwSozF6!gX4KXPkHd+Wv1-s??Ix8lXn~{I^WvqtInLA4W3#z_V;H*00bqqg;5z* zU2kTffW_h2ZkI40p9a<02Yp;wW7dPpG)vz}jiuctlU;qw?sVFb?6Udc?&cj+Jp2d# z_TSg<_cDiA*WEOnR@SGY7PoIo_x9AS=f>sVZt~P#6qt)^k+Ghw1=lU)Fgf2rhx>j9 z&T7i`l6+a;9zc<_mW(xQtM_5*-N-`RnFwwHhqfo`u(13$ZyU_=~t*s zJzY=$T%}lXpY9PhRqHPgbmcpaHT2o%)b%@7FR^`tB{gRUEZLQwsG#CMhi<#9_o7Ca z4sq|Tf13Ns(rmy;W`1o)<#jG*TC9aP{Rr2ufD=kXo`!GbXHG0m_|GbW*Whx@!jBMXtD%{qA5fZ? z+hcvkim0EYW}YpjpH}<5h}rOMN088^aJIl2LHd_6YP~T1%bR~rV@j-P=z=_8{6OE8 z#}4~t5>mE5u1`=_mT=5%)D#U^d~Gofeg&z4HJ6Cb=s}n5#l59IBV!N={uGYnA@+AriZb-0WN#IesyJunJ6Y*dp0V(XSyWIT0lu zL&6&_4cV4Go4M(%%Pp6c()K#PC%B)rt5>aZbr=@9I`|~M|3;VJ({M;Zdg4#V&QjEd z;$4`vNJ;}q*kI_WRd#2-Q?{`>xNhw>sd0)sbGW4@e)D<{qB232{JYseKW^H2zCv`F zX543WUw5z}$*KPM`BxWi_^a(+W~Qsh$vAr>`|qH}53u(kn}WYQWUU}x;S!;cACc*X z|z$=nuWM7Z)Gcm1$+Ar#q<4Q^ZS(+`{@4EdC+yD&-nBF z3o6!MyFY)_NuBpRVxIRhKXSp~r5{(WZ?yX2Z<5OyJyH;FONOZDw;O}k9)UglR}HsJ zFLVAXk8_>q?GfTwJETgBmcx%hycPZzZ^J44*(FoKBC4aWDnc7w^LxEB64l|jxGZ~{ z;t$g^4Vj}IO+r~Umx;wqnOYGdgRc~p7G~Q76$c&|)PBc@Qp!Ny+HLO-vKK)rU+5z9 z80=^6lxo}b^jG$`taGnF`2GFctd#i+K~P`o_orje5_j8r_Rgg>pS(i)bci+5l(GCi3w$cMJU8==nA65wd>yDYV`iS;13!`b%KT;+aJ1@QZZ1aDJvREb z&0_HPYS^nsZmpg-yS?gHV@gnO739(vdD{%Fh{ndoBQ4Wq=yEl{3 z=pC*?`Y(v*yFZ&;*BS{(P+@O9uDIdMKKUUE5H3Fm^@^*}|T zy&gnSZ$|RqwH^Hm8ESa545Q4@MyP%DNfdW>LECBiTt$^Qn_?NY%-;J-kf(?#q>X!jZJVLyo4i9JoS<*!Hx(G9%q>q_uAA zQd{|QnwGdr*y%bgt$>E@J*_cUe5n-n(TJeSb23Oqmv)k!<|(+-XC#H-hS^oBLSa;JJ>x zBQV*fD8Q?RuiXD}k5ut!)KBqGH`$S5_B#sOc+Lv{{z!^@lE6fAkjSARo`e32o~6(2 z_h??@V#1BD?mH1g{e1U`%GY$S7@kj&YHPm1>2_t^mSILpMC(HQ7(M}V^{?GRe|Gm% zdL4_$;#;Ogocunl*>hNCWilcFQ6g4bhr-6g%R8dXeqY~R7Yfz)fve;~Qhcyv_D-nl zASVmo1e;`On$9!xV)yw_b7OP2=-#9A+qQPET}%Ec6F=L2nXm9kdXm5QkB@gDFh6=+ z33oV}@tR#UE?%K$LsrVyI>ctbH+QH#d}_ufdGRW*z|q7pM!J&n=&M@?U+AkFm-{b) z&91Q!O_n$^6h8=v$n_T_Lo*t42hP2E z*PY%ETWjxL+JF?Cw}moZA(DBKHn#Hz8lwtjZ0co=pVjyfMP2Dfya!chyD#(Fq~GLa z>@#@EP%Ko(NoQ};S2^u8TsV{{6J+pZ3mt@wOkWsW*Pt0jNKvjm6Uy+idI9_-sqzd3>dE*jqAiGa0i~o5` zQZ|B3QaoM%7afvgCZARHkyXOHeDu^r)PTPrbKjK`_gB@uK#*t;%X9r0UFEC=WfVfJKiSj->FXCwjj~99(!%G~VT3 zwIc`)j0Xp4O?;)l86(-{j?K^RO89HnW*)lY*e8lR~HPPn1lD`vUJ*1RnQPy1s4M+hv!*hw$42A9grUpUJFnhxm{gX7&yu%HQ>9K5 zjla^x_vO1TuD6zHF$HV1tv0;YMTZb;AY{9d{w$`Q1 z19Z0SjGy%zjg+Qn)Jae7Tt|^HMMU23Kyd3n^9#G1Epd zI0*#kP@Cw6jDI+7*=nWL)G4mwJVY%o6{tBT%jjlllw}Qu9zZU9rAyv7Rx@kkWM-;0 zKWgtu=zpC&AW#8`;0xmi#UpTX%Iqn8pg;?3PM`f*@;OX!sLa%RcHE)$J+*SWSYW^Bj9jvL*D}omUb=oQcIrPDqUj5%CyX=8tZA8aH`M7gfp@RoyXAoiwcy z=r+CS`bkjb>kz7{zf26d9HDB|0QD`KUgLs{jEseuk#ythoLlKXAkUk#Zkichg8Zhz zM>7w(=?6Rv{5e%4t+1!XFpHx0jlyA)XaCD=8ES~C5N=)E&8O@%#}E2h{Pup{K)?6d z&m}P=cNDH~DMK|y99+D$3nN!Y+h33Cv-d$QLw&A)lgk0yQZz$*!6Iah9rQF_#tA}^6_nWE zgyVy_HPv4|UjEl(tP^e-ahRPg~P)wUQpnsqfKW;^mVq+iB* zM?qnYt8fU)yeF&A&wBkQd-q$3^e4NoU&>dOChByXGxDKaTm7d2l1Hz#yV~})*_~{M zGC=ytxJ!;w_wR5qVVhX{5|`lC+ymF=4!vxPQmB>;9C$Fq_bC|4gznkmkPBJWHg*Sv zhCr-tjAz^YtQq!HJ3PIRx8zfK_n>Nt+1NJor_1SaMTg}1H&&1=Lw(^n=g05+qLiZ+ zK)NL;vV>Jf4b~g>t17>pptKIcUH{TY?ltd`P|@)kJZQWjV0ju&+Mi3$R_6n8&$3Sg z3Y||Cg`U~7WmfC{9a3NMTAmxKDIM!BGJr}4+!lnD-A7*-X4H0+RZXO~S@`IdS=H~@ z84f1^1Yx(!pF*n>$~(HDnl!KWGAm0y$h`kFse!qqR~&2Uf8NRJ&uGWsHFxJOOYdK4 z<%=!ILfSiI{Hz=AHyFXj;Z9DW)NKc8KK=Jqvr+_ahNQl0_6K|e{ zI8BHqG3CGUhEIa3-nlHKFB_4sh5d;c+zPJSIEtdEQoL+o>@Q98>-`9~A^1I%i2_c` zH}gVv6ODr88tDE4Zw_z9fJP*W-3m?T*aXqDd9DC1caJLgft=x)?>%!h=jYMQbyO*1 zV#aZK*A4{tPY8vo?H}_<-T}&JZLl{^t9KtioCsIaBH>D;EbgHy0grNP-(0F(i;R9x zwJ3B>QxtmgOpfVcSW_z{!A0ogVkbNmbS!w^QmyXyuuC z<&K7qMxldpMMXpaT_QyU9Nh zzI34?01XWj;WYM|duJyqLUrjyd2_{cSf33Vq>Qs9s986ErQ3osS@Z*7r@e+pl%ZZ) zG9Hw)h%VudEWrf1Ica*Q#ZdJ~B;}X`=P!*L>_c8$k0lJ?$2fD%{@$S0-=;1jq$%7l zdq=V`vPh)*0dKggTjg_=fpn7)IDoE@Qafr-Mo3EufpypiFAvx=nD7MU%HN98H*n6c zGm~WyrcchR98^n@L5BDQgt?Ot7>JEo9JOT#AD4UDgXcyXl)sEx{3}7aWBWDL@H3?_u&@5`UfyMK4&= zaSL?r?KnB>0Zvzk^?37>=+N7!gWxu%RZ*d?)Iz0(1FJ?d5>w*W%}w-5F4#{@*sK{f zT8<%Qq}oKuppPg=1-G%vgT%Fwew9S3aXgNTyL|CpM$FANk2@VX1KWQomizACzdwGL z`UrN_e+c1WmtS?4?fBKVJpLQl-q1 zCEq`cnfpLgezFb$ahYB2sJ3v${f>+?h_+;(qSZ8BQvx3M&E2UW?Lh?F`JFpCdUo$MJQiNn`{}#qaJW6wqX)AzlM?>P?=@+ zlb9U%gxzQGlHE5o(Lhklve+ zJ#2dvzfx+h4Ydl>UNAbi5=ci+a01C-BTs-D3x)&kNu z(!zK?#8IqOgnhS>j(+z9YaPW}!C1=~ZEf=|6AGG%+y9Sx)DX&5iUS_Q}XH1%OnKus}AB zWPFjE)rXT02sPQJ`HoBTN#h#`CeNDfrUMbGB=M3qan|Pw8-5H9_$)tQ7y#D{30pWv zJs@`S3e%nAfFFm@Lf+|e$>KAIk(gt!83F$`2rSV~Zde>boWrdaqe(g;L8?_D7lFog zf92>Lq}Fexe;KD5Ats`jexl^%CFiWleN&LU1%-%_q>b4lQOe$djPN{h1q0k<9zo!I zNk|BqV1g8Kbmmd((T5clOfMzxMRF-;?wgk^m_c#cm@N{f?60z5yw;I*2%W$ZwIi@# zdO13&RH?m4i98c=CS3kacBM(YIdDaJ{`Co5?fOO#;uI85S!?j+bS>|f&n<_tk0T3u zj)&pD#(~@iMU5^xLT1NZn--=o4a+luAX+Ao)RJM`wz`pC$(ehCOp-uVm-&z~K6I6F zXD>qgB1eDxr;mGVJq;~%VCR~iHhRPHwo=Y@5^ADtEK~_K+Cm{={>4DOQ$dsPfOGKnnA}ZimFL z;EVrM)C?0D;f0S9GYaHbYdXr!@MqO4j0%Hgx3^GubPkiTcEr%o$qM1f9YXYMf=+=U!f!zO`0LorV zJ3HtQ(YCbw(+eb?mOz~y>wNomfQP)L`rqd>MKdeDI%(++V*Sd>x} zu?m~|6#+j6*^MF=E%PAHlFi+La)WU=S!M%_!m=iuKX7oN<&Eusb+qGqEJ;AvueYf3 z1=yJazx+U?>Y>-fNFt;)+%OFRJ%Tak5H9ThYrMzEaoafvov$G?_e-MFw>(Eh#T`mX zZlO|^;e9)P;WpLc)dhqQkNj|CjMLT7umu>c@e3cPu)b5TN|N1(HP(b8H0}ERVUX&M zoMZ^W5f|~&0dlP98BA54);6ScuG}}p=$91sbM17XU!3no!!+@ty&VHCXAqcZv7~}>0ZFW&ddS30=wlQ81*#^mB+c+7=Ev`(zX~ED;yUJ_ zIJd0`T{NQ#^FiGGVz|7Lw~S#@Mc!~^ya6bbpN_*-)gmm>+~BJiaurmOLfJAQA|UI< zIiCmmy2BjAknH4vEQIQhi~Kw_P*e8`f4Ui)oG3EVmTj459|sioV2oFd1EzK>dJcKQ zHp2eUAhm5oj?%r{m^Ntt#bDx5Zgyavh+BMx|LPd0`I8G zBuM$a;W3R_6b@IIvxbI>0`P+iV2AWv&B&}x3tRvep54`}wP^<*RG~U0dcjNj^c}JA z|6-r6+xj03xR5}Qlj9nWqLhsnjf|JCEJIyxQaWI+CQ4if>Ooz|!==8>hvAlt6<{w& zo_j^ic5c)VsEb$3k>5gx7p|697c2A*f4>bi#nIx4vipIXcAwmeJaGEfD-QY2v=}=> z{PW)`OQu69)(`lk%wJ7fTudmLc|Z|Q*k|-L!jTK8f2=U^MIzh&YiczVXb%I~t+SbT z-L^x=vYx<$KX9sCS|1&<{<6&`euwf3oy4aYT>@2s!s^e$9`W643)j3k=Kd4v=PrZ# zk7R#EK)GRMd16KF_dAaID5&A809)qr=Wtcv3s~(*FuZCo=MR$AxJwoGM?ux>OiY!` z`VO}9#!vm=-LhU**Kua1%~E<=y5JkTsy7f7brX~fvob(c_0#YI;WY?Pq%o%ZNeHF3 zOeq4G_lS4yjQ|$HcvkWFq1dNFK!Avf8vH8zl_qJL_u#A_Le_Kr<;+M9I*|bdn(^`C z3HRN{+Yl1$xFPVwmVt!P`jp+o{To!>Ji(?)Zo|9W_VP>=4v7@L+@o-{nE_qz^K4{j ziFl>f`VK!&&+GH|*+bi-pXl3uF-Y4?ZSvzL#66kg3Vps}Y(Pphcz*J{l*=+y@sqp; zNh+k%0W#KNR&N_2aa8B_pC?))=`mkDKqLocHKrgRG}T|wD&;R8aYtw$sNSZ z?|*3)=~?H_tp&B-U2g_e9EzozMk`v1MShU}*j`WK#)AIaK2>6-kiBOSfFHAPG-yc?K+%o{ro?Vq)A z*iVa49Nelp7A%n!`FPZ<44!m)O!pp-!h+nujv04^5};#%ygPsi8q6Lk718vAUH-du zeO>P6Ie7O&Le5O9=?aux=YCi4rJWC0oYYcH&T}M=T&9VCRsj2S9$-^-A8kiiC(^E5 zxgSR&caW@%J-Qt7fhV3SZ$K;ufc0aRt@~??Tsj`%9<3NC3aZ8)BZNTZ`fN+}k#IhQSAD!9@`i49Ek@d`;5U%D?uW*F zi8WD$Eiz;Ug;%3dN*~Ly@HdV*&0ysDhOPfGE zUO3agkq)OzC;U7iHPp^v&*ye{^huz^91=NXb;m4HBHG!Fy|d$ii;#w{ZrHAZ>BLHS z>9>b7aUK=5<3v7)tP(a{ZAKq4vUU6%)=JS0L~z=IbQxI^3}S z&}t@7MhA=nK+|uZ&+?vlRO4G%iu}AmW*S1k{S1p@_srClFVPa#29NFE+?BE97lAMb zR4YFR6Jo?R8E^%8KJ~>{qeE~Hd>OkCL%|f&;X75mhfCP?6g-KW0)5u61L%Q%Sw*XO z_=UiOqo~RgB`zTk?ulu&nlGeN)aDr}X=H1$Gr{%y#Dsc^++@OSuJ!YhwTECU8hgJs zNY^X%neBz22gECEK>QRqFZXD%12*Z+Ps3-CC#ePIFPGYbpX0rUpTm!H3=OLdT}Rtb zuPu;&8?-Q*(;&M1o2&3upZ|h3u}0C=GjcN2#fpOmyY=pV z#$jSI{VXNC6E?EdS)*pBohO0Bt8(qzyUe`KFCF2v&vk_1cHVh=oc>AE}z(Pa4km-W+M&+z|Rkux*58qFC_Ij`qi2zq~i}>rj{M>k??r1Q0w|;)$Ds2 z$PK_OO$jy5I)Vz0qw(9bk7GYv#OS?%lih z?Afz-?_L%bmVNv7v9hwVv9YnUvvY88aB_0)-@l)Wi;J6^n}>(zz<~q2yu1ew9z1mD z5FZ~OKR^HB!-oX~1Ox>Ig@lBJg@unCIU*t=a`foYW5{&H6HFb4$4Gj$-5P0s~x%21GU$}5VQ&Uq*OG{f@TSrGnS65e0Pw(Qzi~9Qd z1_lO(hK5E)M#jd*CMG7AE?v5O`SO)3SFT>YYHDg~W@ct?Zhr0BH46(1OG`^DE350* zuUlJN+t}FH+S=Z@apUIAo40P=va_?Zx3|B2`}UnXcN`oX9335gwv|=H~A1e(&Br4-XGdPfsr|FK=&eA0HoIUtbUibpQT+KR-Wze}6C-91su?7#J87 z6a;}lpit<82M-=TeE8_mqsNaQ2L}f~dGaJAB;@JSr=g*tVPRq6;o;AoJ&TBlc>er( zWMt%v7cZisqF%my866!R6B82~8ygoF7at%0>eZ`+goMPz#H6I8&+S}VZIyz7&)Yq?HJ3BkO zy1Kf%yL)oG=xT@hlhtT7|i$Y-+%o0F)}hTIyyQw zHa0#!J~1&dIXO8sH8njwJu@>iJ3Bi!H#a{&kHunfINZ;lKNl7j78e(nmX`2%{POZL zfj}Sm(A1OeRw(6e^WE^3FPt76W_k=~{Tv(FufX`~gIcFWjc1 z6QdJz9zF8i5gQ~gfI!aluemvy@`6DV8bMaMP zRu`YNr9QjWXU*8Hz*zDoh8rab>o9%>=G9-Y77JB@7qaSxc;*k1e6n<(M~e(l7n$Xj zz!!l$Gyyw5Gps>11yB-+4o^H|s!UjZB@gPXM~U22a9Yb{Rk;t&LXemlgT z>3OA_tivN6$syYTssqH~eO+JZXX9Z3xE*Eox?Q~he?>eYe1dNGTi4x=S{9~2mYf$* z?SqN&u=$9eg29D@Un}ovi7aHUA0I4kN47hlVd1WE+A=SGPbs%f%1-j>mYVH#d$+!_ zUW4hlW1Rf0zw(nAmzesPwcV=P=o-lBt5e;hLh2E(&q-X0u7e)cvk*Pf(u;I2nCU2t zpUvo4t(wnLV&9O2WaTY9o0?@_|D}ciy6{BRO_foSthjikKC34j#K2w!bYg-_EU4iYvfn1>uL(=^M)A`h| z2$I>%2*!BejeZMuFnQV1x!!xS(d}YhO<*5Zem0_FKK=>WnPe)i0*}ZT|`znpRUXpFr+eQsq_2GoOtgP8_*Kv-o4SDQo@=EFszuZFW7T_Ea?5 zg_HsE*IHldB8pHv!>Dgt*lf|$0S8C&@VU!{r7wk&9So(A6<%v)lEy?g)lZf88pRmM zJvx+^N*&;fFH4lvri|wrMr8@7CvDmq0EY9;Ig!d9DaE7`Arvv0Wl(fKuA=Y$F~qVT zcD#Q$802f@ZikN_&iU9LynvaXDI6ZVP4zuz8!W=Wce+jW!g}g@09H#ecz$GS+Ba%)<@xMuH0p z#ELx+8fjYxdBhuen6VMqv-)1`vC^vpFbDC3?RETKvOpFm@=&Q=GRAMF15}83uvF1@ZurDJi30Ll zdeYUck*!$cLXpNu>j#Y6tmIb1&a{ujwZG&=<*Q#z7mSU1&|Y}oyefQh{IGV)!c7L6 z!5@AJ@R^b5o~U~^AK|jn>?3q$`91E+EXD)kTATAjB_C$I3Q+4`%^CVpR=F0gnNoQE z41hI-X7pLBnJzHWT!ab|6^5Jg7U&)`Gq@N32$uzXOoN_1yeM{btfIeS|GU6WhW5*K zO$3-#(gTV%BIRuAb@V5TgA@6%{(Fi3V#_zE3zJ|NOJL1-OY(S1`%8Y*9jb3%6_P4P z6zs!}pAw1-#xF+N_3WM?1^@&=7QwG1vuyjpIUfr*&uqCg5S!72V{);*jF0UEW+c zJ}jmr3EYD8jbm?84?8ij)Na0z0bTxzzb^9HqCs^hs?m z(o+)2G7kXMCDGLcg8s2~U-g0nOYkjUpp^rzDXbh$=6#LFkiLs$oX<=zeqx!Kk@iQuo3?5 z?DeU{p9}-iKE-J;eZ*oQPuOn96elcTjk=_FJ$A2SEzV#R+Ph-gTerv2f|lL(FFG0U zD#%?>NLk5&pr1s~D>*nl!fOxNLVPbCd+w?3Mib4{7bx37Rcxo4Ht>IvXY_{Py4-;5 zi*b+W6CDDkpak|cgP&(qZPZHbX1MroBkl-CU89UJ@q<(KD5eyITv;po6gJx`8F$?| z8kcLL30yMJM_m3eKD=j!Q4RwVrAkCScdKB#r3sX{_-*om&G5;k8d8q~wwWdk2gud` zsF*UZqW&?0k2;%iexP{Z-aD}FrYn$u39p!an-Z|GZAbtn#P@#jGiFWfbR1oV_zK17 zQ3&P2!cHt?s9;J}6_JC|e3pjBv@VwSfJ=c6yWc9%taBc@-|0s%gN#ut)>d_Pk+Msr zxw0Zq#N*wPM|GYB0FaMq8gwjX)owd027A6gh`}&I`V9QRmpP9V=Czb{l-Z*3LLVgx z@7XYc&Rz^0>4K&QzFOyhyB_W~Nb}-<&h`m;K;7>N#iIj{vlmm5SR7`a@pW+5xQi}Q z5a2+~4oN-_2=MS+-Jb@XIArS#eDu73R6UAs-v5O({ro}*PEjgqO);wXUT^Iotn+nv9?SZUeRu@qib9*N7HgXnAvN@{AbH56%pBDfE!O4x_o10q{{46Z3{Y zjnZhH>4s}d!QDRiQ#z`jtaC)Hi`HYCeDkV1$z(k?&@f{c@gT&8Gw;v}ww$O^5ChEMD)I>AE0|V7k4mYH z7ZiAobMNI znGju1XLQs@B7b&EGFM0PR2Qasq`FiDuZyIZ*XX=ZOupaMKj>Mmp+Kr@*1ahfLvJYMoB7D7X%K)LgzM@?O1yIHyW-QUn8^N8|$ zl(Pj^7>m0;yP(~ zS&Ua1XkPS#zN#YLKG(n3LT$hrTa>UAGWjgoB*@VWtFC&iB*an+2$qlmj zWsdK0Ou-rS5@Vnzh2cEM#~$fQJ0LG5QXhWQScxAZGN>Jy-HTjp8XQu!?#fM^Pf=$o z&RzDZs~q|mnEzS{=XCY@hAPI-rGKzXd1WO)d`(S#GMQJ9lCZXB=1`D=E%0LS2X^2G z+TF*fC%Ji0ADsN?6DL)pmucqMyJ3!XaT0Tadq;mgBs?N^P$Btl=;wJ>WRXve964oS zi8&+Z9wZHvZtN$vJkoe%2?4QnOkKOhK*K)0$WP1yOJQ}zIi(&mS}BlzaXZ&y^suh` z&+I3>L2mw8<;`H7HS;S>0;E*a0G==wM^s4wfR)m2A85SiE&WDr(1@5uFA;4Z`3Le< zSd~Zxcq_+K2IaMa1hY^sNtAu*D(&0F^^jfxF{Z}^;Eog=L|1O6c5Y3&&cgm zSV#+lqiPGK>u>8@O)+oIUkv~z&4DsUI8D+eb9{u-?@@lVeSRna8=r}c0iEFE5kiTs z|LV$G<3N@ZIy0UZI3(K(cD^;9y8>0{#~IWgg*(2x+cy^yMzpnAQ7{~K<}|65%{@UM zwojtnweQ|b-xXxlr*g+NvvBAl%)*e<#NNVt_ENQLcV=cw? zhzXdlr~Td08)xYE9$#ScHj>S*R2p^CZ*iLuZ?pL=NP}2s>Io57Vxn(XopDoBKyT9Sr$ZbSqFcifOc3VXV02eGRwZf-u}<|I%Q8 zPg|0_UFVrm(Y9C1{?;|ozoyOY;6G`c>qgSXv=#IWOFZ_lGQ8?O?-m921H1>;-mK(twb|OD)fDw?-!59OzJXs5MJEc=0WM1YD z2|+rS;3am>R27MDTQyfUCq%1-@9fgfolqm=(ssOa1(RTPe0vh>0Ui6LZ1zm4f&iDi z#zfm45EL>VYp9Bz8E>x%)s!c01-3)S0^b zJToBTlgBMp`n?jIDDlHbk7l~eq^qYwKV4*$+gpZVc3cmjcb25*W1@E^{ulUHei37d z8`fG;gJ8N)eiVMhiTN{Om9g8B20ZP_isTH{DrM^2wH_2P9L)$|)l(HN>kxzYhDpOk z7|@X#5xV4-h|>Cr7tkl^0vv;rTqZ8{LMXi^Z#O0XW=^Z)8tNuPw20Il0K5-W2b7x4 z0Rl5BOAuwl7FE@0-nLga0&hp9fP^OuIbjraNXzx%ju8b>>P6=QoqO)cRJ=9hK)ono zxV53={BI1F63kiV0w-Y`(YP&~8r*v989t{tJ@X&URWzR)74k`t|Jj^ZFLIoccw5c> z+NC7nZ5?>*Dwzj#VegZZurfWiUWtvZZxqe}FLbg!onPAXwsd6Q+uWSk#`*Sfzcfu1 z>fehA0KnvGY)geMP2Nmrj}~&lN_%oK${n7Eexvz{r9>~t&QhHk)zUz%(uw6lv1?X! zoLD3<${pKkkb9UGxMPqeYQWi<*W987Y2$zijj@o~tgkTeV z0XUyx49qWGW5hQ-Vrwit7gY!|-zwyB2SS1rA7t>a&|NnA5_$Ky%>CfQ}i$;M*~jEv3UMF-cD(+PH<(7!reu;+zL*Pm4Gv=)Y+ z`TASEERM_xF_vd_6FWfa`%$kP|9ECev4N4Kl;a6z6`>tstIYS2Wiyi4^r*V)y552*oE? zasnbkeh+2`L{8yqo_j%X=gEw?+b}KH2Rr<9^^S;!HA+?hr8yEN$S5cEd#Y9(vkEoJ zp|eE-d-f(f#d_u1@3m~?S>_!cM}rW?ZtO<<{rg}4*1>D6ZQ1C0_C$%>|E`8rv88 zWx)! z9hpVRw%PAy_*}eaS;f0X)*ThsFb4J?R|;5H=REJemkt(#-~630&mdve=o$-IQ2|-E z&Sw^deLN7CZB3T0)x2J`eWFlQkerV&|G{BWy^$;P)d2#>0S(SLQ>Cr2U)Y78oD?E~ z8u-I%6y0qovr;MKs|A>5oE;j`xYeuv5GuCIN(>Tu7eaa29rp3^kP?<*+a|wIn<`L? z7J;ztz+XEcJ8$;z^h#8|=B~n@y&O9M6`yf(dc4MSDcV(2ieSe1SJky_|F729hK!vx zeB*a-Cjsc2zyob@A&pX$&Tvyo_30g*09r0(vy%1XLHCa;jp^MZ=Q;ywm!h@-f3JAr z(-J}QO7rFNAq&y01H2_M46sN>Inm#E3H<8&)-wg%?bDf?;pzwOmq3tIqx$#I^YQ;q zpKl$7tSb`MNo|Hwr1Zy`+b0LY#2U)HFnd?i%?%XDT9Ce1s)%BE(9l0>c5j; zK&s#lD)7+HJZifgum6^lq?HQ}8mV$S+vkI06DS?BUs@9K6GhaFv~v-r%)9gVY*VBZ zXu&JB*Gc9b7;NAQ{ed)C_JD(PLcRH4^R>3HgAPODMhu7AJ&u2 zh4Wt_(sav(T57MH*fH5utggGVd)D<*LN6joOr!U#Eg|sYZ*pFDo7iP)U8A|q{m%5e z%MT5WqJ$MfK=lwf<~J89lpb*FQD?dvJBfZZtmGC>Fm+vfwrb@^+%jCNkcwW{ouOcO zMg_#7qII|96xXX&g^o$}Yw{bkBW5*m_glV=9ZYk8A>&oU8$e`*>*Ah6>L-`fGd7lC zd}D&kM4yQX|A5dR`4eGIKrHSkX7oPSlANXcRxNr%8-et)G_p!Nk=4;#x?;P!mb7ae zBOyf7-#X(R+K%8V5U3a*-hvc*u(1&kJ5A%&8T9l6RsA!IRi|n2N!Zg$WBS=ps4v!x zg5JI1tm+REjDeE04=#?p0TntZ>})jrZXiH$=Rj4S*XC*NdGy?N;lmVNq$6&|qi_Zh z;mNSw8xG+O`uvlRb9(4#1_?aAq3zTUtzz&v+SzWJqKhQeW_8CPg25fMJEK8g2qH%qK+;o0>i3d@8hP&b?ju-GKTP&2D?2r{|#GB2hP`_vDsHB!?Y+!2VwrKS9 zYnQyH*GiTG*8G~rDV?{@trDAcwg=@v>4`)hCqYHkT~CO`fm3A&G-=-aG1Fk9FVbMUOkeZDJMapbM{Fd`D;&q)C%tUxlAgiV$7B;TmPf1X9J_qXx#Pd!-%Gh9o}Xy(>TSe1xet(w#YzJ()_k6Mp&kj zmEG6$LEc_i*0+3XLGOR<8aLf#duyyFXliwrZ3Df3`?bH%^Z0$`bIjhCWnlczQ#?PL zHAS*>m(w!awxeHsE=y~|gpZx*q+N7a&*QM-Xx@!epg(OM0`d#gTkf^k0KbG?pDe7r zVz0f)uzvgN6w8T7Yvrm+1BF15w?Z46=QUf?0fSwM#qQmOH}34(5K{W3U<+Bk#lP|0Cb8Sn?P+*C@2Uf+E-vFg+++V~Rjp@RCkf86 z3?0H=Pf@&>zGS?4II!ulIY6L!ibPR-f1DUA&K$OIm9%rSy(uXGATV%4H^iPGC_B8e zxEgs~fzqdzlMD?K?hG`FDi`|`=_g5-+s_>DxVDHKgQyEGqZUnVT*gBp4(kEyzHa#u z2g>MBZ~y3;=X?`qW#LEAVbg-?j3b+IA2ZGxdtvoLs&;Qo(*T%<4RE_wls@Oq@*YRJ z4Su!>c-(*p`pt?bQ0)g$8Cx!ZDHjKd^0ECCwS5~!gYWD3VV-SYI29L+;x;cG%SVLo5UNk#4!ZMy|2K&nlbF%5U-P8RduH?ldeDw`)-T zJ_E$`7im?!_`=q*p#*4L#BZQ|0~4RG)wNy8K^#=~ze@a^*>uCLQm1UE>1fc)dYkx6 z|ET?yE?26O?wCy%zv^~$(BRuLDBfdQ&dH%>VSSIIpW`3DOtiq!Rh5^V8D0tFqqjWp zCoZaD^Gd}7e}otnMXmNSbD)Y)f7IjL1ea6(>Gw) znl#xiB1yS2VePyVmp-@ZSUzo3QH|`+$XaS`({HIYcG`~o2OW1IK_+XVIJCRe65DN& zKW@;x)CmWJJZ#u8x7S(XVnCa@dowD=16{|Pr6;K-?E-tYk-DD$q|TNtY4^VvHTl^F z_x;O$>QEfPCd0<^#B0-u12pR?UCQu6ow4mzm1Yi-c`Ci z?@ky2g`7^vWd99m9H_`d{k|+1C?+-jI#ycpKn$cXV0%yi0CsHg+9!6$k*>Imngg|( zJFC#zWnLPiw5vbl+_k^KoCC!+j7tszX6g3*jpZLe>8W!;4J=bmm#zNUrFMkUn}ofj4I94 zkXW1=oy7)! z^IJ7GrPcQOadgrz!n`SB1jE|uTz{&>oG!<9s5Jl(*T*gf?=*3b5*+k&WH7{jn;1AB zM&VR1;!=4C$LO{R&4)+2k>#BMT;n^E%(=)<8h!e{kUN+9K^bSZ=b?Ry{wz>QlB)b4i0|FBYL!miG;Qva8Jn5?@v{RmB z9Lo=bks)zXt(`m=EFLK0+`f9q2o^Kt@ zYzD}`+(2$zr2U~!XK0@l8E`sP_mw0G200VbAnNY6lTOPH8+TKG2+O~orflhI?T!db z1IV?}A>J-L@WfdFtNxbkt?mxyitfG7^u9bxlC$fb^{K}s{!#s}t1EM3cYFRyAh__; zR`$)KouwQBuyqgT-rM5bxyC>Y;%(Fs&!}HVux6`(Qkamit_TEY6l@&HBe`nIIUls0*e58q$yv?w!y!YyR zAsPTl_AM&*s-)*kJDJ{JF{L*)N( z#Ql4?|63^gpICOiWgKJSS~-wyyy^E&&}#)f`j%Phn^2A=CCbamUrE7S#!2nl-2YF^ zUaDb#^^2*duKMqG{Iw@%KfSaOOuIHxA(c)xm3Ch6zg3Z)&HVQr_Q1(jzb@OV zw4^1^o}AY>-s)ewpkLEx>+QN~4_{x%;>&C$>y5jXvrUtlzHeeC1#m}wNnU;&+S~kb+BD>17O4;UuFkU3 z+5T3A1qkHkJy_&?d-eB>CE>|YHa>CPV@Qg}IrkpO-3p5@F_S`U*O;my(1eZvLqlcw zg|R7o&ihVx9z{ozv6kWPburQz&w=nne=7<@Yruqyb29*;>z__(s@c4y+#M(+PCO@* zx=Ltwe*JW5}QSKn%w)#?nEeC*A8%SH>1g|FW5$C2zoTwZBZuW15 z{Fm~O?sSH=Hs+;?c3Vg9E5H4nAEnsde^b(<^fxu!46M*ycrxO6i#ki#5Zz`Jm~`!% z(icc}_C0v|?pPywm#=S}YKJ84W2Vh$F_eL0e0ZmcA$Vz6D!L{@x3GSLu>B5GJbly| zeOdEvW1fVKdoJ;%gbva}8Z`dlf{V7)BrvXB{<;*a;_ShGDlxI@G%duPN#nN$0I-tQ^=SG`=$t&RB>E`rdp5f91Gd0Px#+}j*_5)9xy+5M4})tW$9R!=GqMvhLre2W+AG+SOjIC`2^@Q>RyW@wM) zzc9fCBYLQ(RX)b#5S^LB|H4O>hDYI(y@WZtVmJB2$hJHUN|~{9MwSz!w}~>9$&@ZP zA=Gl(i4UNrEwzHW!|XvCsD_7)QlriM6E6aLYN5GI&r?Us6Gj%eP?^xt$rV>^p#Eoe zjxX-A&r_jskA;MvW9}TLk-9w^K-%z1LG+gUgP=-5NggD|l4B$~50b*@${t6(6~M@f zYXfbNmp>UkFAP2#6@wrD!TfZz{LmL;p#G=-GXo}^0y*Ozp4<{xF#;+Ton>*I&iM!I^ zny~MKE1xP#j^01&8y{-C#}i6We>-3_)*9OsSqS3x(*rxm}VzU>QMs27b zKQ1t6iaO2#>zVOOQbjH>1YkbOU{XaVCFlln>fty4n zXe;<4h9n@#^6}j+t%Z1K!M96QfeEzP^>Z{cPZ>*f+Ya_NghU(dpFUwaDCE_NIv_%Z zd>s~$G!8wRiD18YfZERPh!eXyVeSt!j-P;l)M@&p!aN%0DI=vK=$We^?T?q5WV0~g z?VbE&A3@D)0^P1hRGn~-Ub2J3@80h;<$adT3Dfy{@+>^ktjQ1iFaywLf_(7Qhy9o# zhtu+t@|8O*PRr5QVMo8KlP6x}J~)ohl=8vVtk^$YIU!!y%r-T`Ceyl}9}X3z-85z( z*vEZy$phioOG!s9@;PprDsH9Tj z>mjL@`()w?k#a|Dcp`hw825ajiOl)2{_aKbSn7lQs4&6;vruMyz zk5p&Gf%f|udfAqoniqaVQ-4c}2t;2plH0F_1UBghxZao!gtKR0^pAQ(Hk$MVUV0yL zm5jDp1IOm&E*bz`Bf|{(hM6Vf!{$b%Yx~SBC8el>w6`#2HPwyNEy)?xsu$PeiDQt^GA5FpDVSb)3p=W8~ZCQ(x^ekHE>#a1Q$9;CaoU zLWC?|!*TY2YA&Op_syTcAJh;T1D|*o)+jxf(^~)z&oC!e$<)|}@b|QDd8-SfsdNJa zL>F9GSgg+a7tBn}_`v#`5rOwaDO9?H0uf1V`yz2Rl=k4XR!Ew!Dk%5)qs+iU#8Jx> z)i>@*3jBhqwqT8z>nA;P2Ki`n>|npvHGv1FC?SC(!eCNF5Mo)SV>TlA@dF+0pW;Z$%yN z7T9CUrYW5ovx^xb4l!l*hcBe-qz_(TItqW<9N%=T*5pP6?o-?d|B5Z7N6aoMqn9-( zGY0)s0&e_tXhe#`B%RYl^LPLd$6>KPGiZ;DXX8FJk)MboX;&@3&JrrRP^#OLe|&Bi zKx>tQ*3cUWnL8O+yDX@wVl$@hV2bV)u|iA*5$eu}$O|Rlwwm4UM-YeC4IjuCK%SE7 zA>2WiT$G0`mS5?16kbln=+N#5(86$WA^Naur`B3{QYY7kGR1r^tf}7OIF8nLn}U zUH7CnYQkr@lSQWxQ#A^|qE09mk&#|lJc6+D!FZVk&NOS-)aoM#jI0EI#z2&>0*lL& zhB`PyV-@GWWwag8fy7ibzga4G!?=ZeA&uRBzRW%!&V^47gC3Q0n%ufO)~ zF=VER#7DmtAknmM>Lr>?lGDIc-iI}2C>y5>9r3iP%YYz%K3vCVW`v}>Db`u1?wgvO zTxaiPU+t~N%FbV5A7&@lTXTxQD+hh(*XrCvtfL3rmZ9Y{t8T-ly+P{2a9g&YQl&M> z?!rSS`Dt@?0?@_rc$JTucMz1Re*w2?@ zj>TR7c6~tFySE0doj&=|VQyI4vT&LazvM?1^TgKpwZJ4me8OZg%spT8D1vPq<@PW4 zGEjkwaK+o_`RUus65pq5TwF9BPxlaMW2+LDg%@58^{ym^&&{Kyi^h}MKczKYRZF@+%M6ax{A#A|m6P_zYV_#Ou zFJ;+1dhLgp6X)Dy(nEz5_E+Xk74A>mPaoDO$LLMWlkdyuyn?yZ>#Hu+Uq@!3J+TH! z>!Atv9khyFvq+=y_=&jXC_?J}aAWf(_+wrZ=NP%8oOVyiTDTN&Dr>AK`X~|? zu$tGG=2FR~NZT%}F8@enG#T@9HkVaz-fyMJXi;e;>&JUhaWiCCO z?-MdCB9LNe!*|@`P9?yQ+C+NN_eLQs7QMz`*AI+qFg9-Xd+X` zhSIpc%=Jq({7l}Jjd!whk=np0URrdS@tAx`nk)mQN+Zc1y!K~9gT7D9*Wxjw1?5Au zuMH3QodpAB;7CJBjHqW=^&JGnPr`_epJqZ=Zj`2jQ_|WHnUULXLpEEWJCdMHU97{( z_N`YKdo9whr8M@0e*ScFy5il<3r$MTFCN2kD~pv*M7h8;DAAR7Q~d*+K~Z3?&?_&I zhHDapeZR&%j6!Io5c>`rZwk1`YKq_F5qs=nu{j-Q7D=;mBE>f91(l!JZ zdVjE?N`vr=#Z*0d~!5a z7h+&IyJrEpr*B`%U@EEvyHEb-AeNeZ@or)t+72jM` zIMDyIL~9z_uIDx(>u}@{sx8v(-KiVZv?uzN!HxcQS@Xr2`!FBCv7NJIOATYHduNG+@3k4;Ag>_X5kZu1M>?}<=E<*l&}M0)zZL= zyYPv7C^@bOqHS$jFzEe!kFC%0%R?L}0}ZCTio)>F5OSbkX-T|+x7jIA3}M=i_E?b? z+)Y$*5^-|ACdf73wISUGAlCs=RdNb6U2`&&5LEJ5X()B=?O)DQFGpr8$A%Rmj>vx*ZC_W-4V{Jl(m%4 z&Cbdl)(f9r_~!SnKZJ@f0D_;ydKDfzf#^s=RJ^sS1Vr?Ytg;lD2ORS>9=dQToA&a~ z9wFqYG`{=oWatI91&sBvGd*NxU8peajhfjkv{tGp6*T#2-SzejzdLW}X^-hF@);5j zo|0J0W1K`!(h~N#IPKeeY<$qS=~kHtO(qfe+rzi_Io@g-pRF|Z9*Dwd yL diff --git a/docs/resources/images/02-intro_files/figure-html/1-pVlxwkzTzFH4PHWzm9cuXbKpzlbSd1TvXEy60Sste0_p.png b/docs/resources/images/02-intro_files/figure-html/1-pVlxwkzTzFH4PHWzm9cuXbKpzlbSd1TvXEy60Sste0_p.png new file mode 100644 index 0000000000000000000000000000000000000000..0758840ab6b526df43385cc3917cf33423a99269 GIT binary patch literal 158114 zcmdqJWn7hA_bt5XF6l;)?hfe`kd~J2ZjeTlE>XI>kq!v~Y3c6nZjpYM_w94ebN-*- zFX#A0e(cS@u4}D1#~fqKxptVcq6```5i$ev_l}!B#6-9H}xXQOb|!_ zL{?Jl%{%?WbcDC+GV@PQnb%9c1IOPpmwapHY-Q_Tq|U@C@O0ytFpbR;OXCpZF=fpv z!(w^&2Dat~wg$pgQiN)W=6U2HhSGH}A3XY7WF+(N=qo%=mE;;>-Ena_PD{4P@Z5g8 z^Rjc%zNZ+awK- z&(F_GN*MG`e;qAgVPUm;U7equ^$!lVc$_SYNJ;&o{?A@<>!OavVD0ko^5W;+?Uz=3 ze|o&T*cquPDtZTQig!(KQ|zp)tZb@O(_yYUKQC|d@!^in`(V2K^wg1-mKFme=JL{Y zzWYo8wfgUFTgMS(9l_1EI&Db1y4FTS$S*ayrSjMbSWgxW4!(A*^quYv!%J218O;(i z>5U%h>w_p^i=HesH^?Wlz>d~J5B-TC5G`K&T@X9g_TwNF{J_A#CikQLt*xz%4Nk|^ z_S4hTgPDqn6n5pttSkx^7W4}6SPzA>D{pmlyzYOWO-@c$S66@h2rDAeW3nP8?$;5FCfo+pB*5{JFWg85b9KSzDX_(j)le$IY#+YODR_UfoVgA`o@? zLZhJXbJne?@0suSa0~VxVJ}0-8_!d~ZGW;rAr<`2?Qj;$lZeMQwDj)!q|NJ!p(Hqc zc2>*4U`3-`hudx*2XQ$^Dguj2x}AoCc-3#c3+b(!n;Yrs_H?7u20E!gyTlKhyM)TW zyY!u$U33A_2Q>k1EYPVpnlv>&9+#9f%7yLVY@^p|wxaF!pvp#1J>|zG+V-HZ@$h30fP+I^XcKl-^1Nq zzsYrPxz$_cl=zt0=r2_G(i#WUV?ml)T72fJ&GedFkx)@voVVo6&1u+XIy#IxgV98_ zOyA*T#eRjloh()pU6Tkw|MuAVL9+2tQ;?fW^@)hY=hxw!1KQ|xp(6!5FR#aJr4gxsD}-=oYfB-O=O}{pdjQN# zg+XBje9k1ufc?qhje`S}AJX;utzMp9UU;~;SfZQ~HUD1sq1`xu2hLo`Wqa^)Z(_xB zCmloNzS8Yr`sic(ghT6<21tF)avisu^DTfWth)7OAmbzOn8zV#`%;Q81IW0o-{|Pf z&p>l!5t_bLR$krNL7*b$wTBhOCf|5LM^^+c`h`}Bg_TvN9s~rZ=-=xKiLr~OLjUgR z>G|d1X1B8n;e$Up2I2Me`0&#G@bv1+Zn`WvCI*A8p`^t6YJVzgnUK>QQm|(Ou!O~S zvGFVviwZL8alFXyx{C=R%1izHIR`3*%;cI*)MIz_IRgW7K?17*Iu)C)iV9SUOvd`U zF~oNuff)t{hE=Z#mc`;f70jy|EWn?~13?w20rubO`AszDI8>uAz?k_uMny<0ke@+8K_KuP#&Wue9a_X7Xs2+Y&5ezaLkb?8hXCLK?EQm2pLnu|8*DZ;4WZMTm7FN_a+K029wrycN@)z zh?J|%)v_1@+~#YoD3-Favcki|Yiqf@+TVhE^19p|TbjFxKqpP%vD*h*93bu25>3o| zTr(mp5zB7U14#!djv#zIRf@kW$jSKyLNs?;?|psD&d$E-CVBy&MyJ6ZjRb-2dUq~Y zy;MWU`xnR?+WT*Q4+27qFm*Wi`2Tyqu-HZA$10EjZk8Fe3G>p1klnmaBrD0}(*~8P z;{yV#e#^U!9yC!$QPF^qkdVYg{BTJ0Ou4Si!BC(7#r6J4i z%mn9OhJuX+Q%gTz4kdKE;MdjFK3%F=F`kTc4AiJ!FZ0^}=2F}~$W{)Yi!F#TmV)?a_0 zL@{Zg;lObHb7^iBjP7d}IeGanpFjT`7%2Bf(Po%E zetl~cY+hv@NRahU(Jl%5Ll;2l2e748s!^V(?^W!1v0Y<61U|IcX#<{RZU;bv;J^qYJw3fymDW==h^J@R*u1v0qaa3~ z9#)@~Hz%4zO2$qzV)=*bJlStMoqs<}55}o)+ruGYPhQZYp|EGs*z!9%==gB9( zs_kqvfe&vUrOotF1sf{zbTcAyH#hQlwDNdVDDtp(F(MKp3F>%Di{PX4^N6d?0~v%- zP-T#jkgS$lZU$J}?(xARKJ0em8{8i`_}>XeblnQ z++G~LS}+Fra*+xZ8`tO!e%o&;kl!I_#O}MJ*%A1x=Pc-W2O%?H$)4!__D{z1<<-^I z@&7y}g-;{c=8+n$GJkE5*dWh!+x%K1A_igoV zSd|g#7YX@?2x$!AuM&ikU}BC|8g=69zgx>s)OU7uMMXwtU|?u+JHX53jJSgB{`+RM zseJiF>1dLhpNthPfXS&9DhfHR!=B6;{vYuJs3WLO>l+)q_KWW}`(gn&Z!oh8)@X5y zp^pB;*{77TBvu-oK4hx*8c)%&{|9)hw5BlsMk;ulI@+26gJ708{aJx%F=F!~6#XjIRc5epmbIwM&mWL)dAr6%?W(B0zM` zbFOfMa0tQrXAyk8m%jVHq@pqz$q-rqDe&F*?s%y=hV-R|h6a3O{fG`HY|6)xDQJUL zj_}1c|IFn;F#vRpfHnaz1~fBWpA*gjpgxCjS19|8Gx0yd zTEY*HjC6H&o(F6bfD@7^c(7-nfZ^c<-*oaqoza$pGyyAE2dG+caq*wUKY00f9f`G} znm`EwG&FY!tG~aW^rgGcN$YjJ?VPjjE&wINMgY&-o11z0`DbShIXU76fk8ii%Do9e1>%8F-0-c=i|-(5 zSQOIuuzW<^4`&Al2Xk_AKvmv&#z7&XYRD-hL@I??p_13w_zu){OLXL(-rgNLGYbpw zhTVCiJWz)}g(Id3xFHWsCK{I4*4CDl9c~Y$oGi98>yAWF9D2v{))W_KqiLSY|1=N!byjc2<&f`W^kow-Vx!hj*`0hZ}Qu`D&! z$OQ;8Si`*LZ{Mzt763Q;22qAk3TjuI%7Wl|{M{=ONV681nMuy_URW%lk;`&43oQSy zlNJBR`x^>*K8zB0p_c*zIeiwHzT5lz0j;F)GP34gggH=3=71a88_Vql zKo#Y|xZO?Dws%mfxaFF?B9AdeVXgFI|%)U%C6++V6P2`Mb3rG(c61Rae|h7*gt zo!uBQtFQ_4NUz!5N(~Wf5giHXeaIh0&25cySZlXXFZKZ*b`;Z`6mc?%-2{`|G?m*% zLtGqs{tj1k1f;yiKX5hOfCc3$CI*FA0TNLrnglBO`PTi-h3GOUh?uZGfNg*v2YRU% z>=+12Py`Uhk5<~qF;RVB-Cy$aN0A8;B_W~!>DDz$F>f_dFjZkd?jvY5{$lkC#Ohu| zSbThZ`}7~(`SK+UT1Pb>LbqE1!TRIl`AubZ2r{9ZKc6H?%j}^aws%-$-i%s*Ij*01`iDLzz+^_anU~Q%7MomZgbxcFb z5t&+oN-N>r5v)Cq=C1(dM?^*KDud~mYxU+O3SBQk(KJG?O4&G3Lg`PEGCU9PgbvTO zS+xe~zGeq@XiQ7=7br7^$YE>;M@MJ@(sAvV0nP`30)XH62K~cQ4i;X)aIt*_i-Ukp z0%>}`CN3^6>R`yhM(v1Z7**2iP5)nU=H^}W#F);{O`rmy#fykn0 zk&3_r`~j*%1fPM03MM)XoIO)?T2unyOpO{~l<3xeN zF2`%yPnr1Hpgi#Y2Q~MYN!o`zR(w4Gmj@~wp_HGW|2-2v%Ui@G>EK+M80hWxi2!pi zNiO~O@lV(OPYv(Rj6azb1IZntYz>qokRzf32nYyljC9zMukHjyo3dJ|zm&jF+i=Tg#Y?NadCIjhVdd_Xc`zdQt*ac%H#8R zn+O@uD`KS}`T#yaResL6Y;oC9q~L4mR-czZ#Q-aT*dye5{(=T#Nu))Xr>IasO8WeH zheO)~9t~2KDi?M;8daa{0^90Sme>c;Rzr{0Z^^J!xzbVMTpC~BB6$@Mkt448T1f7*tWgYDXbP#>VxD2pa>!rBmy9VXh^rFi=U6XMryxFMI*w|Qr z1D8S~4h|>4)q#Tpbe+$Bk?*aWhX(>VOxF9E)jfc$?{PNP*AbzNtfxxULEr?7&C{Z} zH^5lFgS!jpYj1A{%>KQ?8XG(BC!%shB%~@BOG94}grm#A-P#lj=3RaVV+VDj+Iq?( zEKEj$Vh+FO{448R$b#(=R9zo1nN~#+;mG8HKcVGu{;}oCX%*+hbCy-%J^ZD}lO3&k< zg9D*eK9N;J1;Eok6BWT3NrA`(27}U%e0yJL21i7q^`ub~^wCXuiS#m}L78 z8{|*G6=HqccuFSF0WYeKCU^9kkI;am;GfC27MLJWtVnfXB{V486h zcGM9mZNA!U06m(ty(26v%*x6NP(cIQm(d2?wayR>gbdZz zB1Q&=62Qek{r;G#t*yO6l7)p0j3^p~6x*eyspA{hKBW7H(@u;r60g_G5>+`EjWWbL zBbhOCC@Js`Ko|Y7PM|HjkQ-meRKcuML6iGlV(G7;p6{+2w0ceH;Qx0e*v&pt;{$Bw zIGlWcf{+Z(a-F{g9svP{rM0DngM-7;<{aiMH&`8?WmJC$px_wxa|T96u`d;^k2;o? zyCA~>4LqOLbwDHLDa*_IxDQ>DLlMjP>dXd;$l-mAsOhab&ito`;g=_`li1$-;cfvZ zO4LZS9)1@@J=D6&#|&6!13@{ZllRiNhR484<#AZP?+U|i=sx*K#l(g zY6MCH4Flk+WT-z7&53|B+xrn#2)eA$j+srB-Zj&IHx(2`&cDQCOaeu8!X>K>FRLT4 z77^`UN-kZ%E#Sxd_wRvx=sW;BMTJfVTT@U_0L(ik{T6~fP#$54>a7-Nn1ukkgt|I8 ziFJ(S%7EA=6KwUlaR$N|kTwEVX67KQR)pJZW>M_ekf0!Lt8v)Mmw?d$s`0}gllBx} zF)$CrKI|!+0dz6cXe}Lz|0+zS8?aO8pm#>n_&0WULv7E_&e%|F|M-@m%6HTiA#!_p zdw>TN zz#ZRP8wCY9Uk2;LhYG#sgfCwvfZPLxy&emgZxg`CgiKfZ17Z;x8VX2a9i}+FDFKJ+ zt2b{Zea~Ph88ylv$_4rPDpGyqBqWEMo3O9*dJO_tL1$&TOq3ux$A!e>x;|1evK9?I$2#v%(oaU)0_7xuoD;H|JB4#^x6m+w zj}5G%1Dk(sfffCBZpc>u7uxc+w+cB8SxRq*hKK9#Vx#ZYJ(NIV);p|x2TZN9vJ!}s zpArK_fapnhy?XTucTJ z;^Jk1)<6QI)xnr{#Jz(%gsuZ1ksKc%Gzh{fKR>@Ve=Yl+J?W|+=p7Z|_o%1XM92%rZH*z)a9i}HQ4!N34CBdM4FY;+KJYSNF0m4TU{u9iCC~}k{sIR}PPkhV zMGm0&iu#}CV;L>IxWL_;?25)qr2&_fLslOcq58vIdiUG?(!2_8Rc4`SrE>7%NJX>U zEsZBIyAM?Uf3Z5g%YprC^n37-}B5`$uJ$={6i!0P?9*XYJ|jy|cXyG-PzAymeS3%Lcd+uq*3?eVJO=|be`B1YW*b|Nz^Ee*sE@Iix6f0<># zkNvWr#$dh602Y?wzbkgYPC5MzwC4b<0AHbf=nqf-V*!FRocgMF$J=(J0E%Vjj~@VZ zq);E;RAQxlYyh4oaKlDG+s*W$QhX2v1qF0AZsUfVwV0Uy^;LzWVY@tn{$M$1T0S0H zh?J_APJ@oq_DJT_Sugn?V{7HRFZd6FD#Uz=y42$70=fg77Q>g<*O>#iM%*^jWuO~C z^WsGo{$X=Sw^zyEm1CU(&9{zTCnqN$DsC#CZYvBP_BDY71`fk4fPJ>wmX>>v=)cUY zODu1000jlz4-n#Wz$f_*{G~%)*MO^5>!(M=GylP^_R+PqwOzQoW234JhQA9^3cJh1 z!s71g3M(XXz39^FeXS@b2l|`yz+lb{53*o}-UU|4Fn~Q&J!NG_K)p&zN`TiYH+{|2 zMOd-7w+HSWN;LQXyzkzK-@gH61?)db(9;=sRq*H}1bBGx>pq7{&&J7+?j}=SppE|F z_1)e0fDQp}6c4IP^>+)So*+I8Gc!1N_;tJ=q-HC;Ow(%A|E{uce0vwzJ0ag&R~;Q9 z{kFqBw^vH9U%ysTda6d{79+~q8NL#6*+wlo9!`H*JCb_kd$3)e%KwdDOiV00I~#Sh ztWfV?NAWfh_Q!XR6MZ>3IY8Oc($g_W_>@fubzm;R&dAHl?_p7L`%K+i;Xmj>JN3M# zW)*P&UQrA5kQe;ljq#VQl_bT-p8>*P<>~SI4;G@LGO@B&ytlQ6+VzC&0#Q79Z$~>4 zK-cGSvh>GjZH|UYxnc_aH@%)pAO(PtgS9I*XnS-S5xVX`AO{++5T1hz<`NqV?j{4W z3-~A4r3|Qlx2kmBi^4P!nrEriLJ9Z+{#T_2t&g_{K#nmoF~Raok%*fG_)!8#5%T_b zs9~D-PxL^g|KW(TvSE>tWhEum!qG6FT=&NFfm6|7HL-oOlj(nV)BxBINV`xqX7h|B zf~?F?kl`a4f73k56)2oPTn!FxC_@PF*=bN!L!+bXKbWM!RG65U=;`S_orZ|C{9cdv z{{1^3cloU(N+jxp|KoG6pdrAxx(3t{B-g|3RjFR{FBH}mHW1lBIE}?%Jqv!u06a|M zgUORfs6urmd<8$4@M4jK+6EM^g-iK z@K*?uSSdF`x-9Cbk|YQp3#Gr|;{_D}`3W4E2Vj|xgDx)MaoqNczk~4gPl2Y;nqY}` zprl~3ypv^T{Ii)bTnBU5|2oy*^)b3D@<{~WodR*~zY)b7_~C=hJFhnYcel;j$C+qSo+W)ud_tu-u3|)|eAb?;WSF&nO~mNuoOV zd{aI=ORXxV(#2)AmmV2666-SP;JJ;Bpw zQ=B*%5(OvH?YsR2H+}z#<$$#fajK2+^AYW(ggRaEyd25$N-Olox)zrPFYmJd3R#Gf)qd zG`N0FF;L9sa6LMgY?9myCcX38tIX4t3u`cs_@pqa@_kOg7XJOYE^11mV+QL*pIj`9 zYAq>;)ZQe^T)R)Kni&t3VX-(~1I`OgJ7k?E(ahON8+y7_R`iMQ$c@@>NplJ%#nLk2 zhu-vOsQaauk7*YcPsE#P2kpfpHnye(cEiFl3m9;yj~v(!T!3Gj;3Ga?KJ8Hy^U9EPv??Fg(j%nJarYWxaQj?u6>CHCmAml+36auebf09Mgf2FL+ zX9;52j5Q`;J3eN`z#MH0o!1Ig|&e3&-=6imsV5HXg*wEJEzyM|0HsyfbKQoI7$ zhdIFxBBfFOkoZVP(!eD)%g*pwv)rWeNK7++t!WM2*ZnSYBnL)F1?wWDO5xvrnpg$JD#jIC zX9p)qeRNvOv(CMo<5R&+USc?gA{4Dj45Ga)nbMu}*@{V4@U-VqzBSbRS)eu<#wK2a z=%jvUTScX7?ezMs`jx7-mDo$$a+N7F7umvHhDijml__%>%Mdg9p{%{T5TFjG^h_5> z%%V~Z;pf)LsP`>jtuP5|u@vu5>IxL1mgsvb>2PjbnQnutN@&h>`Dj%i>gfq3UMXaKRu^ zqN%(BJohuS4UP}h#Y8qHGRf6WqCG!d^%JY7^M48(j|xhjiMMJnct2Nw3CXvniV?*` z3E9KupvE)XD{o|wC^0d^<2uOU>yr-a^um`NF>6XLIwS~-Gf3n(dA-J&{>gisT}7F9 zMTfMUu9Wig$Vbv#1JnWQcs+mQLB4@`E8{GiyrxZcVvAbZ1nkb=R-NPvI=0qa3dKZ6 zQMxB~@rrO;M@Cog@9`a??r4CFu8yGIRM=@!o+6Qb1|-701gcRpiL2`9=Qw@b;r z=R%KIC@8&sM&l>*c&i0}atPG2m7+pRYL z@f?YY8z@~vo~<||W$j8Qeee9JRqoQF^ZBdPDKAyB|6S%>qcEW^fzlr+{c4JD#ib(A zNFY@v5oabjX;MfeT>dGWwOCTR?&kdfbbKD`-dl@;EMV+>`+g2t8 z7DM!;m0b!AGX<1KN<4(nnw?|`z3x8Ie_5Hb9>m|d{0HrxSF7JZq9>bfcO=2^(ogQlI zz@?C37K*)no~(^3(5M^SCLkxg>BWw?sK&Eyb~0{FwCzly5hmjM>@pH0MH`rXzvb3%i6A-vXCC3?xuBailXzOO6>r14I}q{fsl+erzo?qcM;y@=0RvA$3Fs(M zuB>pCQ4)@Fm@pE9ZcOTlZZ;+v(k2FY74GHPNFnFJ`e@nhV1x*TR54ABYzm1NDH7k& z#N_-DKTNL7nTc7cCF0Im3aNf$K-D3sB>e0~Mx`QJN#YtGVz6w=n5DH~pTf;KARk^y zt}98(^D;Y1Ucl0ol9}%MMZAvcA$n)ZRs|GZFq=t{ukyW&+ZeUiRC0MmlWI*Uasjih z?rZJTT${0w>f8mo4_S*ftj6Y*^!7T+OL@h^#>(ui48|^M{gSCKci2PD)wUZXdu((0 z4j(3?#_N*ob({6S=gVvC^c!g>TBdj8y7T0J882gzg0q7$9#)U6)DmpC8{PAk*1_bt z;_-5m@YB8++U~QB@xfj#O}VEg7Z{9-F2sSgMZG$`x-j^DW~u6TUhHx!VS3ZhXUm#z zIiOw0Sn;-X7@~Ipl;%I^PGt%+S~&E6&V7wY=Q*=LVg)-79 zYioE$ViToaeJ<&?u-ScM;^sCV7?hU}v!8~%ziSnzB3&!Sp%@aL7tWksDjN!qsI8$^ z?W}*7TShNF?)ipiD-yN1GFOOqp@?%9R!kRD(~S8|bA8!&nFu`TC~wG#*?oL!GmE(|yYaC&cb*dmB!G*q_C;CBnbNXI`knp>O) z!d&n8N=`Y_j8NPbUNJo?7>~QBsN=PER#Zk@2f>#~o#LP7k?Y%)FM3`P4_LpR!c7iy z9#!wmzGX?*ahpyO&#gleXkvDcqe@myqQ+p3-)hV@pS3W_4prJOdL}SlYaN56G zIywG=Luy8fOjsuuwYrg)t~*7MP8wm09lfewtI&eTJB78cd8|x@TdW0_KmMsH`Bphu zYHGi{@rBx;Elt(}&qwQ6{&??d>l#%($?L}O$RU|dSbH^winISmF84(3#i)^$k3{h?FZRkn@yWf5zr5nlqigpS5orH z)`XV5iEb2-dPy-dglAP@i#{YoS%{izwZeC)zGV84updwVS#i+Ja07F)l&J`nCKja^ z5z*C8w&$7C_+^icJlVxt9lQdv2?0Vo&nY4S>tgicK8Ge#i_HSz`%1gqVNqq%6|IhK0t zCj1Glr#7Y$2X^xZh?PtWSiYDu+w34?_ooak$Pif~1zj#y7CQgp$i9UZShHb+pg%d%>9kS2FDi z?YD-X_F~G!%|`2V=UFgFp*-#9>WVpa4d|hNAuyj1uB22OnZa2Uz@>{QFjN23u{DV| zy4g>vEx?FRw%ymzk}%bOHugy!amQ?C){FA19Gxup%u^gjf}CARV-Ma9MSed2HTLRfc-fw*SW4ZM zTn62U87i}0*09}B_~0+x;$7Jf>r?aimsaskqehx@W%_e-&k+XiO^3E6u3veR)3Z3B zCfm0@4}!VDmE|`>h%P8!Ci(7rGS+oCp0`*HS6c6Psom@~^b+0{S)q?5UU7dUf?D}j zeLBA)JPlsRB2y2agxzTRTs^fjGHMJydH_{m*mTVb%JPlZipCqNhXIVT`|U@O{J(a` zS1b;wCOP{xWs_6_wyrRC^S&{~Xcza8pH_roS9~SXHbU+{ohCWLOB)JE_2>!ns-5Fi z9`a|s`WqQCUKWhUn>0zg^?$`tr(G%e^;2_dxi@VfaGG5`>hfoqN+Pb$RJKAg;pA^- zU1oiSRSjv;af0Ef5WD^$fzAEF~w8q^=dyKp1d8r2a7j3yh)!P!5oOxsa56SU525z7Fl`a---@RU^*NizknX%au z@VSreQVZj^Xb9@&b9M8+_#mB2GZG=orW)DS6q;Q4Ju2VC)kGINey6wx`*>J}M4p_- zOZw=xJk?{j&T{BiC(^ebfd-xMgy+wC)u2YfuqF#VeLv^`;ey%EP8$`KG2mf1?SX+W z;7EcFYJmLR=aEl;ttFd}uLHiee0(kzMdtlaK9c+gMu!dbTLL@K7Z?T7uX1U9a zUpYl!E8;kw4gacScoS(rxzOcj8?+jHO%`2gkL^U`Foja2i;==-LVb1t8iI^OBnr-@ z(Yf6&SKYZur_nk!MKRRzKqsvX7GtYuJ6!3eD%yB-J>`Zaw$C1qTFzp#*TzmgO(jEo9Oa=`JAV$pj2Rli@Hzev6hP9Klnm z_)7nSM*6KJiBhkg#|!om@Abu*6^Ge+`KNS*b<69Z#+4#O!Z$rr*&<0CsjtGKN0JZQ zs+|_>D_X8(bEowRe-q^^}zTFQU+#|Zk;tZ z7~MTQv<^w? zLw_+VTogI>d{&D&Hg~Kg-b-L7`)WV$YuVMPU6WWQk`k9`-b^cr&zt)kTY(oxXO4-P zaZwXaz1B?ZsF1-XL+cO**+fn0WY*htDLT}by`R3vx@z%b35>C31=khdQ@QYl^t>lU z{RpN(Mc8A=dd-%S9wsUh?+a{soA}ySF{qcE)TSXyxf{ib6+eI zhTN9IWfS1eu-Br6a=YSvFm7OZJO9jUc5ce$=OScqy8RN*2PdNsy_zk*+MLgdeYmYE!Ui0g0|bC zms1X7#VQ9ZnGFZWqeDY3i)D6eGLdaLI>9X3%G#VW{bd_pnd6CQt@Fc&_t@~V%Ma$J zoyLX{S7F@rAZ3~0#}jfcB8g|k!^&ORzKHasQ14i zo8=OyewpZYhzQVR^VpwtaqQGc4~a?lJbf1oMQKd0Gh-9<;(@^??s?sJYYlH7_qONz2tl;1-yAW8#FP2a*NJ)(RqdfA@n4KAtvt^ez`fidk_l??@K0}c0J>o z8u+MC6z$iu$i~95r#H+|UmY5rE!gl;vE^sqry%hUdmZ{V^~ZMmqn$4Iw$PSy)<(*U zA~9Bqwdwh&Gl!ZUcv{Am5mweix+%b71t-KrUr%b3X#nH>HrI(j= z*rJu1Pn^3fwNF=^3@Q8Z`-`SMA&pd~&ttbPIk$Eb&!*{ia$Z`^?}j3~#K*d2n^A{d zf=ljmf+G$nohAkoada*8rWfG^y!t-DPVye^gtyX1Bg@4LJ^?#j%-%WJ)x(xA_qgA| z%_EP3gBhR&4#x$~7dGbPtbZN_6K^L-9m!w>?tO+Ew(|b|R4J}_)@##7pBs)tiU-in z>nrB<&Po6WXQ+a7-lNV0eR>CGPr$Lkv@hB%lvo8|4&yAms%PGmA@sfo3$86@=!H9k zl#C3W_mHznODUcp>raT`F^+OQ7K`-hz>D}6J%^Q;jBb_jM!$S4Q%_X+vdxT!oXq3k zcKw(UrQ_j7&hI(XfxRxX6xlY~uVMQ~S07g_5LUv88dK4{rid1naQC`tm$@{CY>m@0Cs$hZ(sqRf!WlhXinM=n*;deR3 zA1-W8xJh_T`l~jf7YTMBuuLo-9&$(6lVv=pw)ZP|tEYO5To6i#b?f@(MK`;AY;7zf z7fz9$iY5-P<}X`y1M_=y8Z$xI z5GOEl8oET!ppOIN8ubY- zi`jtk4OLSf+CZP2W)!Q`R@ze4Nzy>p82Zwph66gx0f`bl?55kkP+pM9NxqGQ=q%ql zr~I>Ox%aHSYA7aZC{YM?;aG?gDz^K3xw4*o#uJJY0f{UbG^y&YShz+>g0U(-T=e8J zZ*{Obu=zh+G;+lai6o@?bJTRvy%NR`k7aGf$=!?X68sHQ!R0`hKOPZHllulLo@233 z7-hrxDfkpQejWF*-DoyugbqHc`28Y*!TVrKbQV%sfuNZUdbmDMj*CbncS;RY%<=%I zTC6xLE|G@UQ3vav$3!J5$P-t2v_DCB8&S_*N_vgO<_S&)u0s(^urLQ*0}_fJ$h zv8jR`iNDK2SFU5J1kU-IHx)o7Yx673y+)}IF?!2URrRWcmcwtgPj=|#<^Ht45Ii`?7ePj7ggj-ayr+z$gi>-M9yy|!%7r?(l1y;S$-s)ixer) zklZ>r7eTYob6&;~lG}#11`fJSgc~m(rYe+P>klomgovgT7Wt+t!%V~-5m2qqnKW*si%}j{-6L}%p`Dv7evSKMdCrWk(H+2PCvDWp%^G>~)`17N z<=W+RuL4`TM8#O*lO=1&58s+2fsB|*OZ^{5_$?>z{9~Ld!-$EmaqVlkZt|DU2Pfjy z@gw2%?l*4G2#KD?6A}^>lh~j_ZzLfpU&S2(9ve@WX@h|wG;sU~9RB41r%X{v_)fv$ zrV}uZ3yu~VU|WynzP6GlU`1tszbm1_$ps_j+W|6_*5jYdv za(CQpX)X(LPMa{aC0*iWlHI~rJus!PI`T7~5eos6d6zJ98DtNbs(i`E2WHvP19bKE z{S1T{HeX_K6ZS=Gyap#mz~N8&*DDOd zJx}2fF_iB*b2Ui}Bsht+v#}i9rN3xy_WDJ*{?cfgWAynjwyWpKwpAfDszzA?{VlicVo^_^VrPA*o@_B z02VertgVpc#~jq_q$+5%)XttDQaMpk>Bw|`|1CUB&JOaxO3#rstPS%UW@PA$^`1YX zlQa$C7da1}lrk3jqMQ5&amSFna0dbD^gb}WM{NC`);Au5{Zx@g3;i^yKg$pFrstfX)^+Az#OKKNT zKZ}n-8-9>`^6rnsPFV zBqMn2lXGjGXFS7`{*jAq*oqWzjNE#PVH2e69(dmecr?-4fF*D)2%MD+hymZ?0ERWgdh{C|jn;kyy#lq8 z>L(aTygHn_0Otekr%I;4lrI8#yQ($=&1^iNV_Il#dtYRP%kpt|i$I2JkH*_vGL6#p zf(2aNY-Y#%%|2nUMU(MVo1Cztj@zZ(LS?{vDT>RB*v`TW8~%W(7Gk9YA2nW5Qs5tQ zxj7!s&%iXE%y#;=haxqjXF#lB1dQ5>B`~7TZW!4wX6TO`=dyCAwwRjOMM z$dI-0SnWg#ZdkCuiDnK8!lXz_%)1^VKoj#he^xum=0iAz!bBnsMGYl*fm>W5o7|V< z`Gd?AR+sKPOHf0RG6_anvfB*mc)`sHj@kh&LM*2Ca5vDBxG2`*)B5yirzHjS)SQ`C zul@(IG=+uhcILGwC8uq4X|S2E(vhEftRx4v`?lq=Y{z{G^BZ zt(f!;6`f)gk<}iFjirF4-AqcShmnof@j#dlRcX{Cls6sYux4^UN;!zkhR}&nC>64t zbpI~$%*CBJ%>8+!%$M^>{TjCX+tD#w86~$j7Rmfu{U+01%)c#V-;i71T*D@ZgzJhW zC=Twsp7mx#H0;~6pAZr|Oc^b|DUsS&TI=a?8Jnl(+z`X^bp$D8sCntPNs4yjYc2=xT!FJfHlOv#81;0z??N21%VReW$?B-a#l^KmG`t2QnRM^KTpPVwq2Jp= zbfMf_aLf>#AO>UL0zSVhakFBpttMz7T$H0=eifAY0EMw-@Ab3sfHxu{BC~J7v1jm2 zBA?I_l}o@WBrwYdEWTF=)! zdtFZ6(X>07v|;Nr z=^TC~iZlvadG3_!-YGo-@gY%ShJ~tFY(!Xv&U%Ya;`H_&%SWsoPxP6gm+mYU>YI+c z7#NLtx4n}#TPkV#S*SJ^BxLjmMc7=31r&Ar$`@^moDCPz!GM@p>}`_p5os~ti$)6} zlwP)a(aQ>44Qq~mPAbB-5Yi5D-wz4y$THO~rm0UCfjl!(jh zg^K74j5x2~zHcY~!SVJ<=nuCvNB)mSa2jsuPB1CZw-}^~q43;Tsz`LO-;4K-aqjxt zVPNZJ#;qs12fYru!Fkx6ijlk?cr6+%#?HOgu9j657TsyF$%yVqUA6+M-BRC;qwQeW z3A`j?3H=6Nq1%W0_t^svEa5!t+}@~~e2)FE+TzC^yUq*Z*`0Iz6%lk>c*c6la2bNQO-qg>ncFxA`dk=7VRR`d9ksfKuzGtaIA# z8CY5M2XwN7!_PPGO;RNchP>hB48-^2grT&S;K-AC) zuRRHO+BlP0J~IiEpF3o-aarL)=w zg_Bt+vwp)`U6i@g9u&$=V%ko~e}rXoIel0|a;w@Ww=s`AnPP2SS-DSH(m|nSYrA!l`hV;gE**!O;aRx2M*FrJH zYu7e`T^XHR_#a^jcu2wWuC|D|U}H-dvtC_aS7~-V8%63=T#^3G(nX zf=dIiRrr0cN>`YYb8|7Nfk#6|`X72m8S4`CxneJmm%w=6MaOvqt{Bib11~4AQwjy# zbH@cyKvC>DH>x^Qo0t%kjigphrsfNhXg!ekiq6oa(jDoGdk^TG`=JK5*e52d**%r zy(S}2Y5=^8_zeLaIyZuts}r70f`L2gYs*zp{2xuBGF61p1Ic!T3X2Mdm$pfOh^9(` zjhOLdZGj{~pDLNHynhA7Ux3&={=4qc#+0&o99n0-zl?3}W3iyPcmuI>+l{kp`R!+m z5%iW;Ur$e>7=grEd;bH{V)dVhca+%a6Ak9)2Jj{p6}8nOnDu$!f-{p0XT|r+Gh?Tg z{2@cgmG-!gIy>WV0k0S126f*3dr^Ebqrk*5R36*0dwkK;xHX{ToK~{-C)}rQikpv6 zdm3P&*(R2_!94_9vP$#(f$qu~b2(sRv38($HgnxMVn@2az^tmv0ZdU_eB-3P{E22H zWswPaR*RL=<{Kd?R5mB^>mvK|cZk71>!*1@Xj-;xLatO0JrTd_t9O@cuN;L(8tERY zZ}TLwCN&5qrpo4xW7hXRsWUv=q*arYJ0=%hm1In)Zc-n2)?SqzsJ@H2;7$6iCcZ_R z{o#@=R2kP83hR>FP~lUN!n@#NvfXk4-c>K|m$J!XxEj5SSf0(tn8Lw3{cBPgFG6GT zn)^bs8YNUM1czyR-0TUaCx_8j_?P8lY?(Ixuv>B;POiziC^rH6Lj|m@fg|uoA0~>; z?867JQ4RmzZ#|sg`Jv+_Ywr#b*7|37A<#Uls;a;O4D8MdRzbKCztThE{$ilXp1`x4 z=-n2MxTm;yC!p0p0XoC=J2^QhYZwQLDIKFIy|Z3D+4I1Kdn8v!eCYdo@U=(u%XuRPAnW4hR%axGLt4L!a4 z0A}xc9kJkl4=giO0By-_h_5>!vtSoylE;mj>cF^z`n7Sz;mpMhv6Et={&b@DQri(4 z-y{l#Z$$TdJ`jHqjLB8_ylGE`n`~#k`UD_6R4vtHwRgae2xfv!&F4&Dq@+Bp*?Zd| zEpgn^aSHLhGK;3LSlc;IIm`TD(Un1l^pF+dXwNSLJTR6s3sAE$&!e1x%r zaw{wuku8K8eIDj3kPel7hdxgc`+Y=*@HYIi6PYW0{Y|Z?Pgar^3K#>=EjrxZ{+dH; z+_Q^&%0TE8?GEP;Ub87=pToX`>;y#IhOoi@IB`1=flD}!jw`9_Jkjdc;Fp0uk(eWq z#PB@eA!MFgH6!hVmWasB8`O>clOxCWT5v|H{X_y)sWjY5o;+C$SO3>rNi2EZFo~;S zw+*&^pzouRi2VF~yRr66%?Q=iS?TA8apt!m5o9eQS#D6JCc-6a5y-h@w28~WXCi1#$1-yZ$4AB-yd+vM#3B8)VnksEeMmslKnZXh_$Pr`F`h7NE|?Q?kQ z7O_Y9fuW%RH#EFyzI^LRXd7MOCc6izs3l!7j)$m|nC*q|<{y+}QFoM}Q7L0BM!Kx2 zMEEHf_QtXMKM0e3sg5j)6h3u~m-^*gLd$LGN}-k>-%(r4N85aqc4_;jqDvPfYb_bI zCcY<**pz#(kS_O>8<7#2zJqD_oo6GPUT{6-gAl+_L8P=f6rvN@KjZZnCr#cX0@eb~ zmfXf6WPj_Kp=`xRz)l^v-RzN1KHsYUC4kY~C%)V!Qjq6$94Ur*LIvz^NUt%kfR)W; zLW1fwy@_K$4m2$RRW1OH1+PU#A^D1uLHUB67GKP(LMU{wb`7XvpjO74{53YliZwJo z&c?_HT|Mm&Ta6@bZOx$iu$d^9!)^-0IT0#)wD7mG2v%K3hu`6z+kJrI`79Ek4ruA= zgH%>G&1>78{&lmHV#V)+)66mmBWHf)BTAmX;SA=A!`p+bpz90zBxJ$6&oyzq+0YRb zc?ScFxi=9sFOvjEcxYlT)8bvxmGdhlXVdbAjMMUWF+2Q3Ki^}pDCOk-<)1Ja+O?(p zZ;NyLvGJ>8SV1}}cpY^l9G>mKkBB&J?>Iim zayu?oYMfHZl{oD{0-!m7&;!j!X`eST2J5o6*#{94H|xhw2-Dz*EhDX)gm^hx0yQ46^#>Xq>vl(TyWRer-`|~@ zwSa+-^Yad0C;v8E?7ee-+;wQJUUU&xvb2W3r@SXvihD*+RCXm1&7UU_3H?6k@x%>p z?w+W27~GeEu*TcT%_t}usrSybmsY&P4D)8?q;Ty!+}`7b+2CDM5mgGoK@p_bd%@YgiKGw(x=SjsJmm1dbFR;pNu<|&rQJJyl9`V^Ti@vq{PPj^&*TTF- z-x|#D`Ae;9Njr^-Qi5jno2+*8W0(4-1B1`XS{gA5nnhcjykxH@YbvVWJe0G#ATH=} z5iO`^3-2YPwW74|*R+@KA{5}a>}vg}DvF`h zoz>_Sc;{3NR_=nHa$oLp0f>i!-x;bRI!I9>>`#?p7C;aTNV2lBnlu6uz~-NhXP${W zK+Up2hqK0FH~BL^|8MmUtULi)BHTrBNr`crH@3jKm9FmmKl?y~b*=tx5@2xv zPz-`ISS)@9D{b$CWqxiUD*z(J^57sc*sW?epyTk`+{^Fw0s6zpWeq{f!@ zbO>~Kf4e?zSOAp5w}gP5>Lp1(fwi;sFas#8;C3z9bdG6)<9K`;QMP}Wvnxr3LG(|@{^P7Oj}!SB$vVxR{LyUf;A=5x?XOCh0JH-%%C zA~_%mMaa?*_1QpA0L2P&3-lV&9U*E#(M6q95f6;$*!NCO$1?=x%`b8X3&#&aH$(`6 zOXPtwa}LQ)!sReW2~F|2&A9a+S^O}(PEV2R;N+(Zp%Q5c)}TJdR&t^caH}vR z_YuTNp1e;wCpEe2HL9s37Hmqc zeud8L{ME_0!BX4jd3hq`jOmsfAgjK$1LuN6N2HP?@ZoLypS8880h?-{@>{y1`a`*1 zXzO-1I(Q9c{_pe$G%f-nrj_zSR@KM|;?z>rkiv;ii#Oyvx z<0xJ5hRL`xlHKB>7T#(UZ3^0M!_4-FYTY@l*m+{;d1CT2vkqptWcK23CY;edA}3Sy z+^&B`&T1Iwr%AIBpRdDLr&~BPGIEq;6-(AS0QqTcZH<8nsH&?^CtkV!7xDnvAigL@ z^G_0n`nbHb1PE&YjgywKO311S}O1lQ=Q{7fi7JoTq<%GDHg=RZ~ zy+;6E1V0Rx_L&$c_DOi{odI$Kpf%sWpS^*er~zwx9YA^psNTdw|HlQp>UY_Kjeu^@ zYItnoGiTYbr^d9io>F%@83#Ubqfdm&{5a+9^hQCXiXzUM1~2IJ^RNxPZ@+{Y zNhpLeSQAYw+ZX+sl#fYRFgmBT&2a!C8Y+83xYlTMLuC?H`|FPz77~JG{D7QQdPq6M z_@m2b!cM5Br-=Avk2APq{P6I?4bzb`L<`J5tAQdSV*bi!y; zjVE`mVTde&8f=jW8_^!#B%aXf2$BG_Ufl|D*t{;IGyFJF#R~E1z*p=`rGuAiP%BkH zn|ee_Me-92fsGQ@Hv9t{K_=gJcow~L1zISEIeAv z^awHqLQrY7!Na_Eqc{qYkgWp zp~VsIIfI2c-8fGj7AY2mHLj|^ln;0MuSXOz`@OEOF))R(*#4cgQ8l^`IT>e!^{6A_ zuav2tv%F+)NCff6LN(S&=m;~MnxY31k(QWC@b;8Hpdw!Qy}Lqm=|&RAL5$hV>4HEF ziA-E6hHXS@s=xiofT7nn*4eb%b7l^GdDnZ5n3wXkRAwRq?ehXhRJ zJqw!cTsfDeYU!W3#T273xv4<%Wu*+s*Ly_u)kP9i2!D16#eRjf!WXgQksff7$*2<0 zX)N5+a8$L9_d%=jReqx;M0R@?4Z@zoW5G zwQc11ibuP4>+Vf)meXOG)?w?x%zWdG)oHog&F+lI>gAJkD;*1~P`yvq%dXS&HQVzB z+w%q6%SH9`K{ZSOGvKnYJ$?T9?>d0UBBVWny7_6b<4GS7Cv$?H@B9EK#OdYQNg^+Z zq;!!2BJPF^1;YaX)EbH8BIJHYwJ|X$D}mI|CFpDA&V! zI%l4qUDz7{RRBzz3|#u~fCRnyCpc_8kA6uz>><9h{NIjonMMgEmODZYSp&RaLdoEZ zHcBgL0XEVVn0mqHm~fBCfr0~8ulQk}!)H*@CH;#yfwVjl_sgq%m9%8{p9MR3MDx>Qo69Qkxj8x$0 zW*+AMzzCI1oZ^1(cr4cU6D9vO5p~YhQzl+73G-mG-ts_k95nc88FrgY$YXtLo+)I% zQ**_~DR50iW2egZdi#e_fSv0erlwF$)i=j>wu#i{VMRpD+^eHu>H!sfQaSw^5_Ykt zsk6?(i1@&l1#(Oz$c`+3%L*RFdLD%gl``+i5XmVr)a`@Ocn5FGcF~`l)#o8QCm+fs zVYoLGzpk2Y2snfgHP-3B=la^*-W^w?zy+K|$rsX%?(?nvg$mrbDuU!b9h7xR$lv%; zkZ_ppES`=XUoItKXs`O}j}NTo{%C2~*%Y4%S@tw^MB{q}d5uU%ycd|IY^ZPODbi8O z3PmF9|DnkPdB~XP{A5d)34AJIM+74h0iTta+tlHt&+993 z9p4C|>~&ZOjQmG`CJJAU`JMlvg)PAp^g$F{(u2DYbm>nBqd17td`M$ns2`5Q9r>|pQDQH08*bwsUT;`^oi4>#% zkwO#3&z9sBVV-C1fUOSsE~>l@R^>50!L2lu_#V-Fot*j#a7*~3!k=FZuOC+#A)q1e z1`1ei!1X-4@w#R3s(7+k-^X;Z)?}Q$0tRGA8kx0UQcqp6rNSV`l^cRvn)GY$T`84z zAvXbwC;l+c3aM8y)z(N$d^L08TAS4>8dmTDv5s54x~cf5IrW>ULNu+qT=7lXRm1 zMaqvqRw#TtDMcWrVz;-bkfyNu?COsCMKa8eD;dH@i}Qu@U88`93r8 z07MG0uCCqH$wW=WPmCf%)gP|F--|NO>*b;hDrI>3@qCuj zK&r=~3};yOzbPEYASV)!7(`E(deN2%7y6-Sb0TQuB_u}IyGry)Ys=BVWfJ1T`O)Hy zu7$l2RY;)ZcEv2p_l;+bd~~e$x+I)`6*hd*kh4lUT$)@Rr*TN$L_F9!jR{Xd*hf$N zpeo`G=-52_cDcLb!ur@v?~@zPADYgav$0;#wH~^|{VgxAt+M^pnl|G*@^MaMue8X& zVr|WZ4?p*QVuD!uJZHU^U<m{>H%?z!rKj~=PCl*ssF-U3KK=|spm%(jh?+} z=_jPsNFk!{>;G6t(C{jd&qgyVcBB#PP5QaLM^M;pDx}#L$F+uamCob2Vsl;m%UC(z zk;J)bN4-jNf1N7+bptBCM6~!*X-}OMESbl5s;$t`*V(Ot+U6E9!!g&*3sM1#@)A z+^ZnPqt1)?#o{qT5kFmd#;0y+aXg))SH3rr;;#&J+NuQk33Y6KWl!T*{A}Fagp78G z_^`fgc$xfuLQJ&x*>CJPx=~jxl1QQ<`vH9Y@J~{!IfCXhepI+hn+2gl8$XasMs$fd zpVqKbV7}>p&yTpN5m=iXog&*TWvtoQmSXo4!x3%5UaQN`t?Cwz{H-OHNDPo;jUeVC8z z-RyP}5ux~^p#=eg>b^&IM1WdPW6neW636x$xv*o&S_Dbb zF)MB)Pw9iD@^5O#OHC;EAD6{FXkWdZh{EhRb2MO`?j&$Xi%6z(t;V$ zeu~#z5P2G)ovY@Im46MYKjXSW#3aLDb{kBl1qf2_=eylizf+OtasvhcZ}kR6wF(#s z0rF8iA;-qqdJn)X?gLB4Y0F_%X9;5aSvSRLN!v>a?ZkX%Kqvtyw?{qx81jK297c^| zxYEb4s~$|THqe8=EJr54^nwHAve`ch92wwY@2bvydE8ycC;_ClpcrblAkY7fk1%9~ zm}0%A&4TeA*O%v@S;mhAmnujip)+g_+|!` zD2z9#?DWAl;n_L_&|lqdR?WkHp=!>@+OrTlhj|CXrxB~^a9d;YO;~E(e-?E@xq!CP z@;Qi#(<&^dc$qeV(7Z24`M3q?wc~pq1ddLW4|q^v`3g=^W&V^u4o5&D-S2r!=1erT zLp*$iVFt$)Q3#(Flwst0o@nm(YHPd^ar&vv`lJyX%Mz8voxER3AvNjMW*cU0j7aY) z?^{+jrz}O&qDE|^vI(NW*@yNz#3Qth@Kde%mjo`5(7`-S2LD+RsIN5}34P9#IbTh>ZX7BbxpB3@a{P^oD2=lceg6rY#<(81IN-F4B~+X)4mph+C8e$^ka}O z|MFs2%N52DcYDztae-b~{!OMfzcJn5@2Ax?Kkqzl1_Iv|%5r}_4_)*HelhtIxI+(5 z3t1GEEO8cn&BJr%j!!Dzd{URtqAWSgpOD9rKJBuy1(dGF{;OXSAAE28*8i3m6Q7&p zLTh_}B&% zrj6SWiV&aE{h_MqSFjmYR8A0`xKp8vE2aD7_!F7^vmKdS=^h>tlfrkg=cOdYsv}YV zrK_pT#GiT{9PqLd_GIIArp_D%1&lQ8IJ9s(xXd$A zuPirgOJ%7(q)7gWdZS%k^d7|ne+sSj)0UZqLf%vEv~v>>x`j^LE-9YZDS+8>Y_-Ow@F6}dTByb)P=IWZOu zIIYe(zMH!oNAtD2IKq2ugT|DFlK60Yr}(<-+KdSbheY9sn-V7YXQ*%r6rF8>0W8A@ z)heqll5!TuF;*Js=$!6U8wT$8NU3ofL+&Ut*!mG~2xny9l0@imb-5AhXtk__l;6_a zli=KAk~e!B@`(A2!|Fz3=J^-@2zeBWg$a#hkL$H4KNBf@i@2d3xuyq`@g}jm+=R^= z7t?g5pIIY`2})}(l#&=OKhv|6**)+zCF5c#<-K4SX8jNs0U507O<(B(&oucsh5O_@ zOAI9enzgsRZAi`r(jT+k!&BiRwc!Vf07t{V90fzyto`}>MPJcCy?WjsViQj5nF_w5 zZvu7^m@}#L5o)5MOg}m=x7=7$K`@R1yt*AwsI9U+-cwou=gs|IVdAZ{jZ)q}S*ZZf^#I^RSMr1e>rq|wl5JX=YbjV zFcR2OPh0*P&}r}i+@9p(hRd>CH{SNac{nu}QJsbJ(N7*z7!ad{`-g|AQP4n9oe zi1cXme!heL>6~f4l+ZNYV-`=eS5h8RVj%vD$1^EZ2FV9^K$i4|LsR}Y>9;zcQ!?%- z4~FV)bL_yh>b!yKbf*g4MuGP0wD2hQWxqq<>b!Tqm(C(R{* z=;>tq>{RflPbDwNn^p>N0<%ewt^oNCI4*91kplo~%|W~k!0!*zzh(=P`R}oe6Q>YK z_~C4yUkPmPkQ53F$9zU>)kg09&(NYDQy55W^tO-e<|m7shf`fDRx`_|;#*8`1*JQz zWn>6MmWY^AHF~<784be7-VDw%kBFFudjSkZv04JRc}C0qHpiuIZ!ZaaWpp+!?chez z+r2hp@$M1v5455Oy1xfCJ3sfEqUh(uVy)=LvaGzN<5Gp&t5EtpKMMCw46m}J+xnYy z_JrR2lh$h{*djzJ!ul_uImSQh#!novwk%tuf^(M~C2P>_)QyWbQC@P7nA61wpNlcs zC_8Q6`~3ZtnwDZ{BOmPBT)!O*SHarr1JOpP)d1Uw7M-cAt zFD_`orsQe3<8c_|7WsqJi}4&?I{JyAdv$sUhFdKF@rnlA&wwE;0AT|L=#KW^?P(yZ z#QrC%$mEF$$cTM?O_Y#AT7q7TA(~B@C8={)w56vk`2|UU1;TX6xLS6Bj8_u}A8y}W zq~ADJ+)VL_dyg@4VE?Oe8f)cU#;VxphiDt}qKwAOzurTuP`F93-^wGKQA0p0Wa;>( zBz4GpbgGQdIVO^cR$Yp9?TlDv5|cP}qXB9G%L)T;r%w5f@eFF~Fiv4uXR6uqcT$#; z`ufN64YjiUISg(lrE|-573C~_4-|D7Ja*xrKN37tAL($O>Y{@)GMzuFwJy9Y6J$3P zoBef7$&Aj=Oo9qwdcZ0oFy#D8-pqizb)x#AU{dkh2Q0Vf64mg7HpRx~by0;tBmqK8 z&x57kEJ6l?g7#sLj(2J?B*dFH-(}b1)<`4tny4NZS5)4lwFoUaf=Z4PxH}3Nca35yHS2GmwV_ z(b{zbpFRD*&z>MYK(-ggWFBST-I)=qqfoD6&X!L6X>$piuWL&D}WsgtLp`r6I14zy?Nm6~{Xq33Vs3EyaO z(tG3Ly@{BEm(D%=bz{XU*d@;2NG7|F9$cSef;ATvT=B?F_fUe6baa5^&454x*&Q(iFl3la`*#O%zyw^?J1FEy5PvA2h{H>fRghsK)8F{o zeTU#PT$di`TJXs>?k#$mp1A@RHJT}*rP^fyhKo%W_b;95FSd7MQn7!uZphdxp0@dx zU*_DX(EUtLvFnNv2ZwMHaiIje3flbXknE+E*rhz04H_n$O<8vQLQrVp{H%>$_h2it^RGDqnTfX7P-m*MP^*1U|ArLG-UYln`EO zMP+G#!qD@Ark-BR#J_6p~c2X3jrS2&$xNrQVB9EZ4121MvXKN$2;~V1eGtf?iPay&T6P{U^h1@=Wr9ll*>`IAHS}1eky~~J`#`Y$TbeU z+XneN#5}fF)c|b_n6QiWR+t?=)RzOi^b_=7{Rto#)7X~fU{q9+<9h&PV2WSd9kJB8 zAKP$)29e#>aH%I+U|E{KtmYY=AAF6K^muZ-N>hF+)3?wZ{bK~p4yvswfB|ATODTwtMDly zHi1?IYR-bgUXsgX0rI?YSl#gqwny|q=w-8FddI^<$LUP-B?ud72jp?V8%{TH6cn=I zG9MDbb!0J=3g&%30O|I${lRgs+HK<^IsF5Z0E9K>16*du+l)cCRPQ8%BHIE<%c;0; z`=hR(6swX4nYn6dTtCJAu14;E%SyANmc;#YFrOXvw|Gw@G*0SRV@gh-D0h%nGuECT zc^1&F2FsuZhq6)qt%J)bE=aMiI=gHM`YVuM&}xU=dR%!i)w}~?lTC4-nJAGl<|}7c zozvtFm|EWIN3`>MA8EcKEh78+j1Y@ElU6%m8X@A@v?VC8Ssyky?~Y{-`GCv^^p>9=e$7i<#_Ow4YV=9X_z+0V?u zT3uyVvtzBB4@E;*T*u5YyrBoOY8J9>r%~#kWJIgfeVTb*Z-yTlk(^0aOjG4*p+~&s z+0zQvI{sF{Gu2)zVT>P_l*EVlxo`#$-U|BUwS0VkEMUuVUXK&JXcE{ zWihZz%hsp;ue<0&YFLEGNVi<<8>r(hq?M!k8U!W3G_1c$Rx$e(rI2T;cKI4(^evUs zghmCglqoF4sHXfe{mICFwhrf1yoZf)My2H4H!2kjvRzdXqqy^o(>x4#Zpq$;lu*jM zoHtX#5wUomtY}5gv#7S=3esq`I>=h;lTxL`blWK&BSanMoAx8V!uOikHAo|46r$O| zzEPu9Al1uG==}62^^}~S`PCk);b~^0JI+t?#el)_35C$KpPs&>3v0EWTr$6n$(ZFi z9aA#SHzT!Cyz8rB`j2h3@r>Men)I%qCaPiK1zz_Mv(GZ(#v%#vSeN0>efVsinhhg& zsp(Ev`=K0}(NI}Z;+ew#45B^}zxgD~zf;ZZMZ9EbW%`2yRz#XdC$mul2NS3VfO}#Pt#o#r>wUdhTjbKpvV}(wo z?|VmAC1&_g24rvYx<#Yofs^_~rQ3#SOWBQnjw=qA_g_jJTOIx*n8NmmF+40~K&eKO252DYP@3s4c+J=}GGYbOM}z}BLFL5L=Z3_t{~0sGmC zd9b6BNGhOydpzE@RW$dX4&qM_nboA4m#ZtJbz+AVIIL6rG&OW2mBbpYuxM0XB*NOf zG2x=1YRYG|@{3S%RTWNV4h6v%*us zuRM~X*Ca$0rAjZww@QXy<$(^|p$gw3tnYoBA472wH1yG`*dGy&+-lGFGbaykM({; zCFT{8GZ>$1FZ-`PT)rRhB0fpYt-aDho4mriwxcGlobzdFx&C#X-Fj=Kp)B@v1z;Jo zcRT1nt^l>r4bI-bP*$)r_+!QQkg?>;6xap?v*L&xO$nn5b@%_8OBM<$7)t)FgeK+!of%JG9rs!s4uy(elF>S3VbL+(O+8=7CyxZggf1z6#Wo9HxXz`+l!9_h!Biq3U74CpD2D z4>ZKs(L@X3{6-!<@9z}sdzy(sa`V?jCbPxt)G@b4IyJ2iM*I3^0YwthNR{aj9rHJ4 zE8UXiI;&phQDjfD12iP4@Eo~UpvU=qR)0$2Q{mmom~*y`4I9IGUKXB zo|FIxJI_c(Kqm&oC940Ve*FKJ`dehVehz7I zRgmK<^refswist8Y@jBwwr25(ybs1vV8`ONKr9)C=ZMHoSj-Ib<;Va14#>cJhltAf4^*dfhRi73<2}U4#$`w?ly4_y9p=WpboXuZ0_Wqk=JkHTgcy zd%P=)-g!mdC2x(Jp!<@b1F-rIEXo~^7sIQ=E&m;uQ`!mB6FQ{~=P`j@Zjk%|lqHBt z&`Z!OP%z;b{r5;?@0sO@hjKR^i?0xBe1GGEeHVw}f)bfV9jS-o8GQ~4E zq!946js_KrSM2R8@>6m~4p%s)m~gZf_op-~qd=ywzaw)A8@vr0BW;V3xmrhZ*(*8K z^98P_86fLF-Oj?Gzr9HxreZ_+sh4)u4WmR97i>;6ft!O}!}q&WFk8ARqqP?m4Kpt6 zBh&b-TEZOF`*K;$&z2>*TOMqxc&@OyM!|ylZ+}X`j-3eBRk@BIo+=$W|E&bB2&b-c z*XU-F?r+`m@CfRqS9;;Ulcu3e*Mn@AIkZvpFa}^!G)8z1oAzcbzFZW(%_`N}ZX_m6 zrqO&2OS{ryJ^f9dI|V_H-hXmmvU0N^g{rhxmVnUMrRhY+h`mOykrw}MqPcqIx69c#ar%uC0Eve783p8njX>eVD)(=j)Z={|U>> z4nIER_Lnq0AEC?qsc=J``v4AXMD}BSE+5eOKyYmk<@W!cAE~_Nux9;cSK60?At_9e z)PJs#4Yq$JtvVk>hV|=FIPb2cV!xb6TA1@M5mxFU`itD5q`+dH6+w3El3x$XLWj;5 zI*O5BD!m49e`VL;CGU0r#!Inc$*phWZpa^qh*}O~HE(m4DLS^HtrOKiar`BIl5g9q zobrPc51E=hLy}VMdkj+4E5WSELYQktt!TBPXT;75>yoR1$afA_7%oXFV4ENxo69G< zWF4E${iB@NDzdndR;{a9X~Da8>48s$YTuvM>f(98^g6rU`Kg$N!nKrzPPO#p8}(qn z6K zvOg=6A1<$}*zJ{}pjRyakRyc8%BTtC%H;oaE@|3p(R`_Tmk?T?o@U^ym8b#CE?a}? z+72ttK8q0V_b>CWK=@>k-^G&KqM-{gCnSpTxxc$R5zpfy$nJ10N(t7@vwnip>RIvW z2-vN9xthM+i3dS5i0ywPRoB0W=O$hFA-nUp0^STr;I0E}t{{B(3-Brctu|*|(I}>Q z3$BGBhJJ#_FCI>_`RR=^{Tj;5Y$OqXkD4!+FN;*Nd+hj}laNZ<&!k#*1swWq_UafW zJJp8ruzcct4!YF2iw4-8w<}!irpos(zPyzYub>F~P|!UkHTd6qqR4&G*eJ%<^&py| zm}_$-IsBs4u)}CxAvhW$F_(CTe8BgbhX#35%q8yLVxmke~-;KTAVARDOUq+^k zJ;Gf0!Vb2*J|l4GtyKh5u;50&*;6jeZitclEnifIr$z+~sjcX6dGhdpvHKD%x*%Mx zu*5oV*0s7A zEfW~hY(Od~GI@iJ(SQ6EESd~=Sop3F2p^RCtC!%d7tb^HJ8pzU#5hAey|^f0q#FG_ zI}3EEKmhj$hr0&?2~YtYKrST6Eknd4e+02cFH|q;zZ8v>HUDeHpN9*a){|&h_7&W? zEPS_=I}k0K-@_dY4J5}iTj=TWtQ4DhGzIE?m^Mdv1B*dbHp9ayuRT#72fOwS$xitO#)ArKOz2wQglSWZKIK78!$pO!a|S+o)_#@mTC)2;45%^ zp)xpxyf+HH--t}8!l&(@snX+AR0xmMx`H4_hC)YtBMJxJ%p7(0^CU|uoYI-9z3R$C z37FrcU1xm6S{)`Phk_*tXizoGDfQn8p{BAl9Bc*qC9*^6Vfwkxuznk)XDLVYd216! z!c1|;s^uscafxOekcDxeM#$nSbX@H*8op7r*77IFUZG~|8Inh^`jhtO+V?RW;hny* zt5)Ltu)@&yY0Tm#;-=GEX-%fE)!Cqy5?fo?uSOvp=kxhmBSc7A- zGT!b94h{Oa!^?(ZC`dDY5n&4)!=;<}7i|LoNomZw>fn!>np#e}nEby+5cu?r{m=XM zr%MQ4?XmNivO0D=UElvbT=hMyl*oDcw$=w8oejqS>(T?GRf)Jf9cX#(N2!!qW(UjC z)%9p-35yshQll^}_+@DArEn^bzsYaJA&-bicSd+q)=?_urM#01rP{|T#XCe(;(--@ z4J85}ha!WHz#r)`L>2ifHMRz&x7uj3ijbsoW|vkjpBD=yna?yHilo3c(uf}R68kLS zk}6EMKJU|Ae0HFo4;RmrwFL<&!$~Aw^ryrUW=v!ZY}xa0sop|weqR2@1;)uZTStX3 zmS2uWa_UQ_a^=RsTAM#$>se_|_Vn=3Pgd-Wq3f>iT*@ZGIL}?1- zY7sCa*=uTYn6#U!5_N(3tLk4|)|M^LF+@0&r_#3_lF}-is`Sz{8h_r3)25W;GwosT z0$mgV&8I{@b~v-~h}-iZB)Re;ipwSq%5mlTyaC>hBJhK1?rsF|{sdlL!=B0IDkOVf zOq2QJQNw^vA`MK2?KUcSvs-eeO1B+uZdPhsroEnQy|QWL0?+#hz)f}xyp%X9!jm8a z5ZdQX1Fzi)*#ZPngIWL0R>qIge_lo~;{_;>pCD!!q&Ynt)sJ`kZ<7kL_38er7U1FT z&n7n)uj!QjY_r=?Qn^%pT2kruEtM)Y-GT}Wy|l1r7r=6ejw_ZS?Wd8|ueV>;flqGP zo9P`vH}hjP4P{8NIWbR~HMXm!G~wXLyukd;cAF2E@wtIRQ@MYl5O!4PJG>@dX7hfw z!4AA?-;L`Y7B5OkF9Ex9FctKQz0^|jt(*m$`K!OF)B_o0@YI~d6h?S&^Wu|ixS50P zLpJC%Y%Vwza3#egQM|7i)wE*D1Vtn#3t2l|-jk!7D#Ldui{$xw=g=v93j7#mX|&P+ zH3!dMXE9#ltU=4A93W{&IkXltY~V#_>^#XKi2;uosS-_v&8{!_)?gD4iH+uK)85qI zp6uFz<}Y#KmhHXZU-25am(Ln9b0(SX*;zblO@;#JF|mk`QLJLM$WPrxg2;G%t)|)pc##v28a=W7}$M+cp~8 zYHZuK?WVD9v$30`-|F?;@Anh3>labO?!Z=-buA(qhEd;ld0&;!DJG=4>kSfJPd zha=z)gC$Mp?f;Tt{x6d5826oE<2*00HTxVVx>QZxB$kMLkU8|;W=nG0qgbIiTb&u} zLo%L+k5_u6AB?_38Pl0;d03iZOFkJH9{N| z7^VzM>$r==8n_{Cx$x)aR@6)SZdoe>8P-$(cXU!173a;d>9>i%4Qoj7&^|E1pXRiG zXpv|*D#ygi0*x;|3!9eq-`RD%`qDZXm&vE`>;y*mF%)`GNkK@59MPvD$8*MYUG@p1 zGlex(5jvUaRv%Q2gY4l?=oovtH?|NnTdy8p`d|15e?wbt|BjU*`*;6z^l($wBRU`5{H zW1NA=UV6Cr5C1(C&UXxn_O_0LSYCM`RkdN?We4m?d#6`mOik;?kotE+C5H%ICy1ty zF-_UEHjQXOpN-zd7Ykw4Hu`d@OCpl^8w)oPhsp4SC{39v3AI#2SEAvD77QJbfd4G8 zMi0Cr3>PlctFJVEEwv-rqKF-0;E6~WZgnO|T%J)kO`gM6u`y;ak^$LbVEp0*ngGfn zO8f*v@|78Tj<0PTS}16f?jnxW}I87EU$tiYZ}Wfk{4jSeMtV1_9}2!-GrT_)Y?5SmebB zcj2n%;u?a7pebn9_w=mi!eETY&=8eg<;pJl%ade}DA5cf4xbjJo4#F~7`N3bz)B_Y z72Q;PDw}4J=#4%lw zyz5+D@Y{_2yd#8;^?e!vutOVvQ3H_%15U)S2HMSVd<(9W@b5K3GT6Efrbb% zwC`}Ruw+Sxb{-IzNvfo2v>bN+uTv2ugaZ7Q!$8tMKF^7jyQ@>GG z%r{lQX6c&8hhc>Q2}%Ra+XrSt;P9{j0*;)UPl|P;2Aa*S(of%vYV)VWAtN};m;)I> zlBiS-&zjLVidws2EA`>OvGr9qDm6syrl;=^hPw(z3T1w!8xmjga+BbG!mgJZ3K*AU zQgN|N%!r!SJ5&1+7YB=YqDC035~XG>ImyWW`&RX*L^iTd{r9bOYWQP1T39=G*waSU zkj8S^I5D~!rQF9!?as61+_)9NKZEz{2HY+Wi!GQfYF$XaZtwoG`5EOOspAgq$43e( zRL8)VBA0WI`6cM{X@mIRX2~dq&eLetX%FB!68wBQ;0e9|{|8a@Q+jW%@LQ3#L`(L2 zsmg_66AVf@!BmQbw)iKXu1q9;sYhHS>!r>^$%S2(wmF#4L$xtyj5l0XHvUe<-uu}g zHOxDe?Pop?2gtw4ZLDxDgZC(Kus`}0Nhc^UYET8tGt}O+k>6CISvOpN6yGze=Ro!q z*~Sa?>0VhBrF7C}5R%>TyHF@>p&&;1 zirUo~P#I}T&6(yY`}nHR+wGy}3OvP^yaiq_@BcCb`JkVvX@ZVp^1-w_$~4v0)!l%F za=udMzhT#wlAP=VBvy6=X&61ZN&fplUh2ZSe=-ORqRxl;GpH-a9+P<6qjfAlO6IaU z<*Np;s!aO8VmFYir%rcr{1oDSC=j1&RAxVcg{lcS*y)9nmtf9R?dxzoCIkiI1f+-gb8hOpt36*Lj^`}|TIcr8c7Ijpcy-0jM#I=G|?3{>a;pu|Nf_?Cy?fUzlmS_YK8eqVZ6!#g%|{BmT>4Q^B=X6+t~ax?6bOG0RnxW~=%i#xfqO&i{FttBBK%BdJ3!Otup+Urv2G3|7UvO_<@&T3 zXFrK>9d)gikRrlMwdbDlR<^X_8ixghX{oJxuYU+j%OneQsEu6S%Qpj|#LJzcOeA*j zr4*HK?J*~qWO~!FwFcAD95H#_NyS4w0lbHM<4CMvB=F-%E>W?!s7PypPw!GUrYS39 zH28oN7fmwWh{J(ou^)rcdk$#?KCiwRE)Oz?8unbA+-yj&!{;gLPc zi&PdE;-17B9_ShELLk_W@|LhwRwn7yORx+VZg&o6kW*0VpH2g4`pWRl-sXCxzj81kJW4a;M)m`F~O0EcKG2Sej=fRUiXT&8mDaKRy4scnOE2v4JE@6lbofR`1qR(yowacY;gYXGxfx_)9Qbl-)o)W)X7KWf5F4Izi_-%yPMKuCCG!p)#=e7bZRd+PVHE zM4ehAMb~{xzgx`&U2Htuj;73>=EJM-Jb@amtnU`=R>zHiCNt_U$5r&r{Ju4gQr-X$ z;yllHFUDQq{qvnYtBo+AGF#~H4cx*h5C#W)twtcdFPF)N%Wl001i_n}JjW&?00)vG zav#R!oqUjN#!6fH+?D{_N@pJDh_BYI+%xKkyJ2i?kWV@%vEL)P$iVV@U&*AbICPFO zf_xUK=@UU-;D6S-U&PIvCoM-48FGmG_UQF`U-XihJ7Vawn5ighdnOb&x&oy`16rsp zWS+I~Z#;w*1^r^fVhL&uQpN;TW*nzX<`k*ENKnDUrOubDoTE?C6rT!t!+u{Pa!wN% zc+#(6uwhF;uhR$0OAA)Eys~|Q_BezKjtL1P+6e(32@!nrP^B=?DC(;{*Fmw$ZnE24 zLV$Y`sQ--Es6Mu3?4487iBv~kLQvH7!^HzPHXH*FB3PIqSB!9bwrr`_eBE+SqVyZy z&UMczQ)RKae>*c$w#VN2cq1Y7!E)nZnCWb5YvI#U8%5s6Q{-_C@3X(z;#*~p<8@v; z``9q8?_TEk>M!sa{my*c7M)gOG0AVUW)JE6CJZxl8ZpqgS*pIE$_f;O%|*$hdQ%`l~=~;6Xo(ZBb$}eak+GKxb~zSueTvTc5#jh9HT@>6}#}W z*mzfFhCEjE1ZCZ8|AC!!t`%e>VZ=j+pbgoiki~*;vrqsU5)io9)Qmv37$(PEr~#Ql z^Dg~^CLvIt?r0VV24{2TmJULyzaC1K@?hZ3@bDJ*)X3!ED*@-PuLMv2m`zA&n8i9j z2^3s9$CQ;Pb}{3!T?fTO)#hulA7PA9^=bts3M68@Sy^j`6R( z4htt7{EPr7(9)k9j#JZLS5)r%)KVX_3;1|82G8ao1g zA1P(*b{{~iz&SMy3Y56|$BfXIG1v#>x|?7EU~Rb9%U<}!Qq>CJHv)Bs<7;5BZul<> zO6>!LN)N_x<0%KmKLiYZ7Mp$4!r|Y*u2voG38OaK$qMr(H6#YGde79d zXiK-juc$uxJuRuH>G{umouSjz8vuja_ySgM?z=1$U!F(fV3Zi(0g4wwMRJtI2-D7$ zf(8#3%z?{N_ezJi-wu8aHHaEx$3s92DpmDJHok=zCf=VxY$uwl3hUbV2bZ=XTu!Gi z;7|d7gf_;EOTCfiAvu$xB^;0id*6?Toj3FyKWciwT7*@_CXJF{$ikQFn@rQT7UfT( zNbBY+R$%(IhU#VnwDNTc7E|@`=C8M)IRd`nvzD0jxX=HJ4#H*`*(?>&DLhar-GSAfycrwUk_qr^k#>v8-jrEDX zh|qgP76C4#-FkBdkP7{$$77BI+OOq*+HW6tk82#ciJZeYS$xg60=7{2uL+BmuL^fD z6k>G{nXk#35Q6GbF5?A9uHZ$B;~*Ax-}LAEe8OZqS_|C~E2sU37iA)SkoKG4!Hj$e8JcD0({=#Q{x}lPe(6E2;3IW?p*C?&ej zI3WZurBHIU_8!QJ1mJKK$aAji&A8NRq5|H`U_iyav(0VVNy2v6_sfA3N+a2}?_j{M zjKgZ_@yUQeSC6`Dg#}Uwzh|}kqaP^JwX{F1vu3qj1Fd~_vs+liPok5%81_D|^P`2# z7OVBfgMbzBQ5E17gZPq*e)>g!&m#YoI`(KH?VnU1D1wZ~W(hdLp?#k@Y2DfV|GiQD zpz$EIol=gw$t0Bf{)9WdP2Sp1J*39b|6bF|Gf4Q>kagk7{U{kkse>T&=#4I|1&7u; z$Z9`uEZn)(el)~>)OFD3b*3eSDkE5ROV$`bUMp-)R6L?3U!UA+%m^H+7NJl5BT?cCWxPz!C3atiS1Ut_2rQ$CiKMqoqiQ?$XKUZ7tS1_@82OMZ$5bcD zx5wutp;N(<&J^l+n1UQ-CcbzDyKOf^p!h)v-RtSq)S9Nd_H~oUVf%@h5g*0ir!KAm zn~UZ4cq4E~)%rh$0u3ks@lCbcXLxraIG5L>J#@zImvg{>?D{+Rd9RTIT}f|~`}bN@ zUv-7(_0mH=-#;KUZ+{h7fYvl4;C-B7@pL+80tDC%d>_@+8Gymg@xS{7?#o`h>9Mka z1))smJ1kWEvAld|rZHig5ZF@&*H)>tv&H4!gRS~-r$bhODV?Gc<`hvkzKtFEK>UNc zrhni{CAvil0kP4X+(*bzi4cPvESQ~FDtKyU=*L$n1b*!&nTp~mu;(AaesGfDzmiT#Xc$cp z6CGIy#=q{8T1`+_1kt$0hXoi-EZ{9t#1SJq^D&ZT_nfYVgpG)8fTH9~6N=Tcdzxp4 zsv$UH*-;Y(p4;A5NRzK=3g0BqF5YSBp;e1Igh@3Sf*@eaUdeWs!XdCUJ>c3$aN2!X z4u@p)SuVYHyJ_@t*%IvFXQjvPrx1h~;4)V%xVrN#y6*z{$QPM#lTwCuU`xPxj_0;( zn*f*N&w1h@r2SC(YI-0mjoiW%_^-kN5D0+fV{5H;8vucfK){E~RfySxGM3Ng_+FFHbr?x>{UsC+pbW;EEV-BNV2n@@! zWrBcooJx-=w6;$hCr%R;4TLBxB&M4PdBI^CmMCRWQ);LM-Lu5zhLm(p2c|5N&-E-@ z&KXAx%`Pe{eL{7P0xS%40G0gxg}NscBsYlLQ=||MRj3b5LW5kC19jS6X;8#=KS=DA z!`fJq+>VZ(QXtwOT#XSV8ZtS`rc8xAVK4mxMeK+`)fuMfS5)pIwa(u#T}GadVtJzd zXf&D1X%t1S=R-%(%mSk7X^KAAM(oI z5~^ffE6R@C@$~tm#(ty1X6T;5?=nMUS>vi^wvG`V0p7%xPOHsqvsn!Ac5OHBjN)~h zTPs%%7_)u1$peF+KYJ=&%DP`Nn4-k++R1kUKlf7Ov4$r94a=oj_SS%{4geq8>^88x z8vngrVS&UVLu4X~qV+qezqOI-+qF^zd z$BHme7;2#Kcs6%{Vg;2+m=cG*z=qQJN2;QRr)u~BcVro{)&6iin*BaADad4e8x51o zljjfQps`_&ZOoy1M~)0FXT_SVxG=409zLa?bL;{T2>RJr5dGf+AHnAsC!i`9&4kcK zb2y$f@cl?CniR{YgU;2^)E<-5S_ULu_CJC;hN%>($vELqWf*yI%cbXEexazBq$Equ zV$?uVW}y1dCG-VeGid5L{-#IH2K?HoXTDC`A$3-05JfR1IFXmfWZ)PlVm{k+(_pil zf9sX_!>EgH5K?yA*0Gy;E%Y}4UJ%w7P*OiX{^U80Q~ygJc6&H}Vo2ymmi@EBrKR&1 zrsSsyQrXF`gLciGa2he35UA%UA#6{^=G_WsEfyw$)#qda!Pa3(L`oyAe0{fhxod{R zIKd^dE>=7AqJ6X#(8+4IWSk@-5S%!e0+0^ZiF}| z0&RBfRZ3xUr@uvw=V$|0U`)nEd%9I8^NY%u;6N6IHdW^rq|f3I_w$#jy>PMy4RdMY zo6n1oFrZxr-nWj-15bGwm2gXP!I}(%{Xn1;)Z9_RxmGsMynE%GW?YsTpe``xe!{c#NeQ*MIO z&O3i5Adl19agDtVu)T9%9a+@tSh`M6@ll)7Qnmq3!{apHbz+rQbVjG)L24WVnj8eD zhw3N!;Fs@4U>^XxBo?2u7U1Up8^M^3hYLi)UqWYJg9$vj_x?M(S>nDEZ&)EEnSSLb z?oDN6Z@DQzNTZ`gp71&@e)E)(#fd)l9{YHgBW%kFbD5~XF!bSBYyz24M{6jD#xHzu z{gYVpXGML{zIvNdupqmk&{Ge+Wr9s!nmZd54wc21iNe_sfr< z#!k`TW#JqS3BGJd2)hEPh#IAs%co+l^azAa4K+s$`G{5w&&7O4l_}#9LS-Fa34&du z8mcQ)$pJ+vM7i5kWY0*ktm#Z|pFy4~s`|>&*B@4DcfTjU71+w4tKyzco;5mI&jDI7?!b?IxDLub&H(Y@o5^HpD%N4;e9lbI zom&!CTF%1p9i@Q}YyAz}vw$m3N=mjM_{Uc&BlCaCPlxjf(1+*0`}GyrO7izO%b`*D z-yBT!UnEB|-{q-0Z<#YtS9G=DnKEk^n-d}AIild3dY;x_upEG7sQ^a;qP5FN<_20nXq{FApL`OvT*p0))bJE;6qwEVA`YP zv58{%bz49B3tQA~M`2b}Q*A#a$>14yIEDo$8XMk*I7j9BUdFF1-V3(?Owy#jx^n9W z@hJE0k6>{VU_v&b0;rlkzy$X9K=88&h=0hWF?^g?)!qF*>BbQG&r%Zx`^{QqPA8>G z4E;-Fp89Mvku*w6O+J3^+BvJ3meAPZHJS3t{_7NS!q#xeFhIMLe2w!T|FcI4odYDg zHojbP`kQr_dD%+}xY@PG^O*IXqQr;<21KJ}Lw+z-$_}#W9M=xE)vaQArv?6oK1bs5 zk`zI+wv1wdRZ%K4ZNDMIO4^m#AaiUYSuaOKFqc_lP0lJK%gb@+*lVYYLt~-zqs}5j z1m>TGk%(6cadVnI)oKoswBM*T3j%N@spo%5+6bmM_e%9tDZl?>|o9mR{;Ih z4U2qsn;M1-_o8~$lggeCTi|8594ET@fU)eKf0O#>?R^yYMBim($^;@e3)Kr7Xcz@0hXWq*JL0x@Q(?B-S;8}*t|#3K z{+=J@-PzCYH)dUXx@<>I=@)>i%V)H9P#8(|SCOnZfMkcQLV5W7tV|XrX`mMs*Hyt? zIA|fai%2gI3fLo?DwHbyj>%AtgAm6c>`2m8TdY!45H&UD6-sl*oSUCp$PvNU zM%#p)Q^U+if#>r>RCbCPB*z!MDduK%Y`yU13$-L{uoV}7yENdYGL#ux9WCDs|MzY3I`oqz*RTQMF@Or>f0!U)ZEfv$pS5!c$e)zm{|lVyL^xZ% zCD_Vscqe+0d7C+>FoDalDy4G<_jH@d$#@r*X>S{rc*g!wD}U@{wr-= zlGlgID}LfE}i^Q|e@Uht1##%@?W9i0sz8}m={2M!o`EV>te^@`a{Ks%rt~|%eeCrB$sZ7??QYxzjA*rEtho!rCqGc^Lf8$W43hg;VYrWoQ zJCzAr<-gYk-*^@7HO7!!+_wYGnLk_tj22%*7vrl|!hobUH78cM>8-4vF6MK}>bx=OJjJQnk6hI7n}PE} z7DQ&a5DFQsTm3p&vv;@fPwa>k|GPsuZWQO{r_jzDA|u~Lb`+k$Y{6a!_CMxd)UtZY zb^MATcna~NRppxZ!@u!#HdM<=G8RBhSOVt>5a8SKI6(XqOR2Wh2cmxT|Gm^Ng%%V% zWeOYL>7{<1$ji_$P<9+b8%c%m08tyLjb1VB`MpGe*JKlbmv8a}_4QsB(iYZ3?))9$ zK?GOwKmv{Ev^0uRZ?oMWZh z7Y8q=?wVzqiBXj4R|?+_Ya(czrBi0hN_{Ag1Eq4&?ILwSRKSqVocGS9DQ35>eQiI- zS(<7YvS9fS8?<{2A|Wir1dJkuz|=rsiNN_u!y3G#y)UaNQE(6SDZqs^6E+@0m~>|q z77loqB&^F+gx<{F5OR4gkV6%!DyjOFj`l*9=ZM?L+7sYpmcB!Vg-CFjGt%Dj#R=~5 zYCMF~=#+W=(y!N^&kL#bnwg`2$8*J0!GTKzB9gDeBZDb~jsi5oXwLUGMix0&iqYp5 zED#NHV3iKw1Fb~#u%ym0a-_<5bZ^8dsEC9zJK!$$+$<`J+MvBT-%N1;A&zNMKLNV=CXIW>F zA4>>Ny}f4)aNYAsviFq3)(EXxetzu>8ZIF38tjM>5t4#R!BqwlSAG=?^{uB-S@2kr zh5>=hqx(c9^-}o(a#h*Vh4f9Z&FgiQoBU8@9Rrn`>C8=N<(U3jHLbss-Df+Fhe}FA zdmzE43`mm+Pjei18iae3QyJ(BEBR=cN#F=+*g*Vpl8n>c9TWqm=({#K}qmVK0oVSsgb#;7bWi@>MJ zL9m0QSOXyi8BSZ@{m>O1eJ?v8gTQ< zHt+U=mN45kS56iwYV3PY=IXVBWRI>MsF6mraDiuC(tBCA|4rzZU?gg%2%!qpD|LUpJOw+`S zAVC1aDNO>J-c@vS83NmY6v#UWN%)EIDlRuzuIO5nwo0-oN(E5~N5;QTm>oG3dK%yZ zo<-&8AsKdL#({o8N+FUT58FIXQ2{HT!0EpEU}QAA0^O5UT9c@6!U^1?;(*PysbV^1 zN*gBZvjT%s)FR><-|IkWWa0U=A|{^mh@r^mw8|A=oy^>Cpzm}p(@6xAei?x!o6)Zr zl}-q~YLqnE3)xhz+@BF8-y>g&7M{`^jYyUX?fWK($~Nr7+*K(`x*;}%UiX<0e6?Px-1y+jbHjQOt>)10yxQ&h zBA!5@lwbQyiF}9;5=gmz#Aw002d`Ia?EU}4cM9bw$g;iFD*{WqWr`R6}djaL0YBH8eh@^{U{2NLnu*x&)!rFGXnp-Owdu*01fgt3^9voYk6q>!FCj?3N|B<7VD7&6VOZBr)s%(!W#Q0>se z6zrHLkA_rtblR@$yDkMvIizy-;q^%NUrZA`Yb+>`?8jJTcN>8Xu}*{Kmd#m<4U`qD z5GMEv{2a0viI+z?ESPoSX$cs$#?@Se#f$WvO`f_hI5+JQw@=K%Rp3#xXdPr2ZdbAG z5Xmq;3Bo}$GGq#^6!cNQ&LYnTB%Pneoi z(Me;=1W{yKjfME(?`Lxr2}D91` zTu_6iCLVA%QdN*rjL?frCDYU|<1RB`nhnuGA*mUfI)<`Z{V=X_7`k6F6?Z%x^tlcv zV)Cv$%CGwuGFt`Oeq#oQD`;ydO&~vo@R;0&*}K2rMH@8tToh1Z*H#q!Yt16R-LUT) z1Gn`T1z0T>*#|&jexZ&9c-4mrj)zCX|B)aZA>LTuNHE8Ovj1|SQczFL7n5PG-8W5E zTFr!=2IopPSBEP&#$~dS+qbKdNy=w`(zm+6jd8J?TGYVV?TfpOMgiXqff%?gDhPb? z&Do)w3*Q-sTNMRyQo!fYF71;-F-!5GwJ3y0F?|(OX{l zW%sY>KYQEFND+v}fqbdHWS~d0$^%0#rIQ2|%IhKhaKZu;cR|^?K0oTs=6F=M4OH{Z zre2D}jGKrM4q<-9r0zCkq!6Q+i>5%s$=6s8X+6n!ohs@pWs}O5_NSIxoA6H?2d;Nk z;*U^-a0UzUs!l7;o3PqA{xDt(10NdP<~>PJv3IU)YhmM%b#FHSO5F=7oLd_J69wQ_ zd)ax?3qlc|-tF$8&)E2z4LLJ@E~x~k*_B6pJTO+EECIO`AQf@9{_7c_X7&_#8H9H4 zL#-Ao;rdIG`y)Rf_uX*ZU-!lskXKh$cHU3(0yePw1bthzUKht-t?U2yN#qJSEIkX2 zst!afn2An(VQ6E_k|;q8r35PM;cHklW;evJCMP)p>Q1!dn_%R85u}0H`J#O zpXXxF6@90N^Wa1p7oDfFg7pJfbzcwyAFn%3X>Rj7yeM10Tg__yybjlxkA!RXSo@jD zOXRa(?Yu5HYtcXds3WX!)*1&eLp_(|o^zI~rykoL3%?NZof7Ur^GRDjEezqK$%0cP z-m(Hh{w;=I~tODStgev+B~* z;lm?&m2J#KM3p);S=m(;^*S#VpecPtE1E48$vaofSZ#1JG_*5TQW#J{7ui5oSW8zu zT){{j#0rh5Pdc0y9z!5vDOl<-N#2?!k0HU;1g9edMpTSRFXM~SWGz{2cpM$u>8+yb zgA1ey$f7a7u;X#q1ZgqyxX0~TcK(d7T6Nv_oe?_;1dY2`+}%a7`PU#&_`8C~X5|a$nSeNkmVUe6q87*ZcsDa+Ki7GH zcsqr^^arP4m8Si4IMOmzr1$iW@brt6(G!9BD<_a#4}yf}0cyFxa}?P2X1~+>k>$|) z57g)+{9kz;1I1DvXTIJ=SS$V>x*U8u5~MspA!OVYU|^U}$f|PF7f`@aGYZ2_2MLHf z!$llH2AE#`x|3Q>hc-zaMqq&u0YMHSraFhQ320LPX25nvQ8qv2(Nv|FwELZY7CbhH zh;ha`MxBP_;Ku-VL%CZ_GDxtz++u)*t7gAk$ru-iZKwn_QyY!y&%L%)9<)*}*ZB`R z6JiZ2iv{@rGXm{fJsu-${`X?WQt`=GOwxw=*;}m`vUla8((FHhDrH|mVR1oAm-o6< z_Kj!}?0yL0xxJeD12|3iSg*=KH2;#g&;=uX8-jJ~Vbk+0%s*2%Ba`~mYQ9YN*NSKI(CTcCZ(N#ew2J0$3dNF%2$fyahzCun^M^n4 zMos?C9sSl5^WQFx2D8;-dZbba`FIh8W=wX(P%fm(^Bd>j&<`qz5O#cFw-BQ;#U3;$ zt7C}ZL97olsep#5F%h4yb47}vP(|7FYk^pX#60TalB$SIIETW z2ZOg?p0m#A&GlKC`paARac(#swau3mtx#CtV#MLi;nOmR`P}BT5R6C{h7u!p226C) z74>PXRD1lciNB&AZmL^|6{whnAGaVCC^)88@BYLD=rkavw&X0@%mw!CK>h@HZGvH5 z15q4)^YNhRl^svHxZDEfy)^JarnAn=QEKbB*5a^vi}0~mvjo&UqT8$;|Z<9#PQ*hD(NziWvriO0Lu8) z2T5DC2wqd?lvw|v3l9#N09E}FCS9eDhRWq4>0pWvsCg~M4ZrVeVVP2_slGtM{^z-G zeeRlEb0sI)eomb&^O8e2jkxs@dZlI)A5_gIAqZTvQKuNJJqE5RL5b=59(6i#4cJc= zJ6#vJ)j;?7Qx&A8__S~)N!AHU_MTG+Bf{gw)DGggBn68(5i|qSWCk>5Uyrdj`k_&| zFVCNYfb90g`*;ax%3Aw;3%4*J>ZmnVXr#l8t7fS5xln8U5E=otC2f8*jOPV@Hekvp z1EjU9*@B@N;71K0u+4#h$EyO`@li|T^XjG zWdAENXO7Ew2)SLC!{FKXw29vfXAfa$T`7}zukkgkg82{%?^208ZWBIk(A@b$T%vy9 zGvyI@EOC>fB?6O`yx$w39Z-@;3I{CXjff}XRmsX(w#XDvztXn8*3ixx@2md^5be9A zbdX%u-<1-PQ0M{NhPEdkzD>GYg0-*`L@5U*6sqAJ1%5R;v2~-Es&)JD_uk&10ZmkN zgMFFk@WCEeWxx{?FupJp1;GY0s^s*3M0w) z(r}dR>Tw_+YN_!<4s+#(YFY75yexbON8iSKineo{_sCbW*NfHSyuV~MbvOG+ygNF4 zD?HZ3r$q(_e8Q!p{pstV_uA_Pf%TK3C+yJjVsB`LV>TvnGWiaDABw({@0hJK> zuAV6xBMR-HRozljM7=idbplcy!^Y#_EVIQyS$8P^L4$2|57-S+Kx72T$};Q{Nv~Zk zIc;DSg2hyAqkBKabSF+D8v56wh;)y`Xt?@^PN9ttcj2-y;c>5bl*;W?C0;KPx>Yy*wEe@adk z2Y@;<;SQXh$m{*j`94NP8k7BHEyFaCXDW(zgKu@( zP(y8MDOb2{Tqwu;tsV!BSMA%3ggDsNGeG^~nUDLR30y{i^Y4I$TZ=SQe6EntYD8!j znT7wB9wN{7L+MJxe~px;YJ?gIM%6J6-7Snj$|+sR%)=}?nz|y1XgC@}E|0@=D}GOCgs1s(9g8EtVml0B%a3fTI@RHR^9aUabWR@ftaR z@h}rCJ*NlbeaE8GZLa3CI!NO2%&%A)GcMCD`<4QQU?pf+k9}2${wm#rx)d0xvH|PI zGa{(RunW*`ec|NY|Mp{PRo@iebBP7Nl2#s8pU)9JT(Gb{SFpz5AmNI<=(aW-1Diqw z?z5ldv~{1}tBL@5V9!95lQCPH2n_-((C@QlZfe8DYJI;kP5(<@YjlD;efVpO=jRwd3yt>kwJrGigMes@aU_H^|FWY+<_%+4h#YuZglgWSbGxX+66nj)qIphd* z@5}bJP&v>chmQ0Z>ZdFBYks`fa?M~E^^1+thI?bwtXv`%5Zb(9@u?%xrr4J#S?t%H zEl%fK16Bus@%QrkQ{}f^KB9-r-(qR!b04-Tuh;^KRU-kS=Awc1DIlBE^%l0_#cnx& z{sjGJO0%2fq2Dqf;h!l^x1+oUWXf1)_?&;@TDV+q^Zyk~>CTS-Pek^E?liBPBbG*6 zgQhw3@5*e0YiyCps1!xh*~B?G6s6#MowfF2${vpITO$o2%sZa6E7$!hR>CVQn~y#J zmEi9e?`u%qrG%2W6$cxU-|}+5uNLKj#{ZRkJ^vFIE7B^(NwzZ@_X(9#ESM+?aZpG+ z3RnorH{zgp5HqsmTog2J%Ax3{(a#y}Pyem5nX_Jn-4+YG=-(?#kLSI;Wi#BqZx6f% z&leldQ(j(>JUy%Bm9)wmdb*_Ap6e4%M_5&++IE?{7?Gb8=<|=e+|R$!=mz7Op0z}b z#AJUIR8Aios}s#RO=zzw1=q&RxjbvhSWjP~(fx?oWhu5nCc8_YGyFOwiYc8pug7?a z9?UfO66O<>6fC!J@$Lxq4s+;07-aN{CZ@Zj{)Y;alqXdh~FcGiY(${ro=aKLPvwiA zMMOs}ADN+j%IVrlb^MLSwpCE)9<%q|%yvB-w0`kjk}O2NAi#?THs&k=s8s-Y(a_KU zylEMX%LT27A?MvQz(C$2B-KEg(X%%v6O~D0|Aj8k|QGLO}iJ*(s=c4 zL?dl-mdaS%+`K7d?@Y1`El6%y`j;tWr&rIiIdU_;ie8B5)QRSm>pHX<$IB=E1Bu}l@>gr;~re;vY}zJ&P?A`ZhDG91ezil>9x@Hrq;mi{;aT$-e}H8 z?RB@Npxr`~m>lJ#X4+;QpHy{j^7>@k=^LgHE0~UiA{(_im^9)D?kO8sn!m5UI4vT#{#dJCv=%L9!wNUu^@$EL zE4UUy#7IK}2M6Z`G$bvivpx3)L;D$Gg}f%ILGtr{sG1=^d7^9o5HVL|sRUP(-){EYq_&6IAPW%=R2%Jggytq07Pt4LE&bhUHmSA{il zvmstgE#Hz_#K%x(F-}2Te2EMjm1)0d_UpRpbKSmaW5IXx1BKeic@lN_S zgY|HfYEKOh#H*I;w6GF89SYx=9N z>hvCyjx)iF$v?i+?&la~9>gYbJdp() z$SEZ$H^_Q-grW(&LUz;K$ZOOZ;vt3-m8|ksI0VOWK7$A3ud&Y5o})#y*&zZ+o)Ojk zYxjX~(@+lR$GBgTPs3qAGe_kN?!7pp?#X;|_7O`^Wlf*bV zP)GHcF??G{v(G}Bg9-}hNH)UY$z9=3|USPUy&JGe!cwZ04@NrdhPW2v26MmIb8|T(H2_FJ`QCpXI z_zLpm*f}PCf%sHG+rdrjZQI#qk-Xesif+{&r4RM3gQpTq)y+aZf>zhsfgC}jUHr0X zU3e|@d-JVjMD1soxHad>hfEwYkbM8l2#|+=SYkp#Jz#g>ioQ2J4b3@_NDKro06%jq z1qtvYRZhXQMAF6+laej@QNjX=kM2YlW(cN`&Q_rM6@;@pLk9u-GSwh9O6^Aso|@|Z z_rszXwU&T=*Iu1TQ*Z&dyfeg7BrK_my^k^hMz>#r6DChD3Z-)amhbQeposyk*cY>p5&po z^>te(d2Kdz9u970C(=f#i>cME93!`6&LF`Wv{+e8Q4sFf&obEF*S6152Ke7ZgtRAoP0E@VIn@%4l}zJ zH8*7A5~VpC5=9tSkW?{+Occ}hj&TN>mCjiLC9on`vBSbpYAP;91FK#t_F6R-iIGg5 zi=t7yf15a|U0B&TZ`g@7pf$9F&XZQrwEcffy#-rT4Y&n5ba!_N($XD6E2(siq@;9r zcT0D7NjFG04&9~FErT@N{hf2qbMFse9%lA_bFFu+HBt>C#gmK-gEz>1zI_G1mim7q zRRc3X_ix%Cd_g-Acd0IRLs#s>xioW;#(uHdd|kLIp*Zl^2U3=9n4V;KG~v)$K% zH7-A1t>j=SWCTLdBDmi=I#crJ6NU<{f42-~t2IV5{Z54S_q(y&xrnc_X7PLWUsZFw zZy2A$!ZSXGGPMr%Q;zKsl1C9{Z+<_vZtPG>2_C0JS|MBjw<16i!^0EqZ6@V9i$fAu zTXh+Prs=e2%55!G_H{$Fc$5*Lj%bra%Egx2`_;5`9ma}UlLNA9ybzv=R_mHVC2KeL z?wF{|ju4Vnom#-U`#p`Dk}vo!r_jy5jZRx6Tj?P7gY%s=HMr^}?Vo;L~4I(xD(>T-Ga0B2b3P?;tIpYD{ zDnv25L~4WoDJoi<1G|W5)JOg z(@8`uM)aA?uN{(h5NpjRR!?q?MXsph8)aW;uo@gyE5hD27x}cZg}?Vq3!fcqLpJZ_ z>Ak?9xTUwtd>V{18m{d%*gv^JRC|<*zm8tqp%F%I%hkp_zREK@7pG0XtoY~O=QN#Z zCzeo2!z^!FgN8g8@x*MKdg$c=oG@D)ty}HvwYivQpD^o5_vE?XfTOnXKWu zu{%&tl-%^x<*Rn|)C#f-x#wRV_9UwjcZoxdY>3=(vN$y357nrR<{O9#24%YjOvyf^ zg{m%x`s9c1oNQl(GrFF9%YUnyqnpsfT;ueGd)er07wh)Yx9!DD=bjcecxZQid+q(U zWCu=HA0Ur$!j4aTil;{C=z;q?Lni9rZ)k7uL+qe#(C-8iQB$}|h5Ntt=2}gZwKs** z4022cva6{h_$>B9WmsJ^M1!ES(+j_}w367KhAPV8 zRAbKRiZG26W5zvBpWbK*^|DenL9^gLxFAm1-e?tUv5t`V?;XE|Gy9EGmcN7|X_;<_ zT;;!a#-y@#8y_thD&2%?r;rw4X07^5!=Y_g&YR3NOXxfNvxX{S0iZzipV=PCXU{_5 zEw{F|4uVhy!B+n7=U##v4+<2q_hQJ5`pk4X4soMPS|>Qj8Jk!|gwR0)fAG`?H;0xF z2~Vaep3Yfoag|7@$P)5D*{+^0Re9Srk|;b_?m7s!2lL%keB%grzmIIC>{vZ*m6dfw zBf!O?$fu}K#B6kZ7;to45XjFeU-vkFVAC;Y;+$ykh%+rSjHDu88e-iUc=t?9ogg%r zChSSLucl^!RB>O_XVk%+yG+5@m7^8KTwk5t5htZDU-9XuD_~$}6yv$7bwrC8zQt?C z&G)3WK5av{ay1`Hi+tk|`niA8+uNM+F z(4)UaE*|j~58<)dHybp^3tu<5#VpkFygy%Ze&-2va~((`k6Y`TZ-1V1sGc#+ZR+sE#R4O0`PbM^nXYKex12K=(M+O_)T=6{|aZ+9_+S^WuloOn+d}$Lho2A zCB+Mu`w7%1kyQnKJ0CQPr^-LcP_N)qxymI|z07cgtjI=(Kq$H1jD>R!<$XmYslL(~ zPkSj0(~b^c$~oJ0kZm`&_}y+7d?dkwH)e}9+1A~tJjfAma%sa~vZsr3>b)bn4)w34 zSIsi(`As$>Iti}8#i5WH%y1z7x&ij1eooY1|Qv5H)PAuGX7n( z#xrs(k1=K33P^lfC~l1*S0A4JC?*wEUf2^djJX(g+yFf`fze07pDPi#RPc6;2)H z2}s;B|C6}I5y6if*BKRM@DxTLQMrdiFtdM*%*M;a7n)1pPa?f@HcY{o99IYQP6my!;HCjm*Obk5F-7RK^%~maW&}T?;yX+?9pfXut{94qq1{>R~hziz}lGuD& zIsrU=^u07j$FSqCDOW2{?%>5t*JNdz_X3L$q#8njT)VQx?Tcn&7LWR<+JI_^slp-I z>e6W<%h4$&W3%5<^QIYGr5x;sT1TnQ?C{%bHs~g&;5sMCgU_M>^@(nXL@m2P1*3D6 zEt^j0ipF`?SKqm930am}Ugzn8rpMN>+6oid%L0uH<}2knWr87-8ll=KyOzfztr5-J z&IntwUkt2dc{zGA2g~8v7uN7fguKzTt>FWsNPB{2x0JT?c9*>MOUJl;zRUfvf#`Kc zlIg)<#)qB_VDGSGzv_1>8zeFkUz)xDU2h|8g8xJ4YY9otcDgew|AF#shbYjH7lQ&{ z{wtkV<&w|Vz%=|*^3^wnm;;+y><$n@L)0!nIN~nJrwklvhgbCje8j26J{$o9!R`5G zm)Et${r#lyFbSRo4+e|T7;|@E3})}gwrEa^@wcgt!VvRJgRHHk9&pRTGQ35poYjL; zn%5O0y&08qS#*`5`O&fFrv(HyM_Q(oO2u$u*W;-cdhDFi;~x+=u$Hd*(`D2(afF?4 zp3XV4DnV1}5Za1>Gn?Ygg5+KKeoCo%YMJ?cyv2RIc?#t?dp}a@&b3t;Yk7_RF54GQ zO<@fUR_{&&>5V66MfXa&N$vcsyb@>&%d&ZRVImd(TnoJP1r(Kx#mu3hwdk8clD@3X z@2s`+WiW7E)A+K4>hI3Q!4p@c5B_0FnZDSN!+WSVFNIm$kxlqjx?JHcB^j6Z!va$E zylSo%)iKC4WQ8&|a!btXfl^6_$}-EWxp7Tj!%@L=tjd>PL^KW`p%Z3G@cV7}mQkXS zLVHccFT69!h0sB=;ia*KrX{b~>PVq%npM$%oKfJ5Lu?)+ampNghUG7ZdkIpZFV3Q4 zXDZ>7`)Ryn8Xyt=wy!|1^t12t>sk2Ol00J$Yfsr$+Ca!{60acjKK#|5J}0UH1A=%& zP27=y@oCN%iuioiTS)ig7Roj$m;v1v(DS_O0j9g-RIqn9nm`|*i2x|D{L&-kB^76^niUXeqJmg4NV{5R#{QIKG^oAX8qfHud}pGCWWK!FSjBys!@`R^Ka^6 zmNFe?bqdW{3}*?^^F#SRq4^ImM>6i&axzpp8OC0E6)Rmn)rubLX#(C_s(@SLK)QOHN^a@A_o@ijKPGwPE`6e zcbdlQALGAV+O~aOv*E1P__K3!d%jGj0hV6F!^67~=@S1X20*Oge5D~oK>zM`+9<{9t_hc64Z{VEcIyz2t(Ye}~i_uwFojtrKau z-!03>4J`9weRxMtZwhcpZI|o*1E9+Q97ux98lY8ww3gGo`?i=AqpgtHJz>bcw0u0@ z(W0B`Aoarlzg5mi)Z>21bOU{z-bicwm?KAAD-wV2jpm|0QH$f=ei$q@g?m}nx@umh zHUzICzKhh{{->+zgW_>!MX#3ma_hD-*v8N)I?`QDJH*P`;6 zYYX-gYq=$dB?w?&Ezh-`#@6T>+-nTaMsfLtW3rhQeK*uou9vDocbZ-Nq^ zzkmC*8ppY@Y(sRdd9=7Pl=3N=WCqgebRGXEPT$TtrFpnQ0r_ez$Ms)p(j+AN{St?% zkGcg(BZ>WWGX1YH_RlU2il31T^D9}GFAL@I;BJajPpormIXToRe-+HZb>iB~n%$Fb z*0@d-gvzGWCwwxql)ZcsRuZKA$(G(`C)y=ywRGGTCPNhLd|Noh`7AH_tT;X6Cx0%S zrQtgBGCTCDBy`Paae`9NV^leGNZ}!JBmrKokb0Z~wD&!) zFOP+Vh5h~gS65ennY_x}5t!D`@&kM9&%?tEKrNL<%=5D6cI)~g&moW14OnK;(b4@^ ztIpP1ZUOb&Sp}L|j-tZaIpFPsw0_mo)m;H9)Q3`VpJX^9!{82^c1?!3rMrc88Ei6v z_Bj;9hgAripvcL~oAQK`aXr^M-k`mQB1UUzYUXf?B&V6Zn!iT-D!-P_ysjThLzL31 z0~Q%Y_p&lFTqKy5u?bk(0vQEv+UW^VAh{DjT5^bzx!6xsj1$e9+ob z;E5icMr{CP{cawo4%`FThR^=6jD$r+G`*eC$A zt@trnWTU6Aur?+Pj}LG(T}pfRd-Vv8TypPqM=?U0v@$!qH1Y9Tql-z6Fa#%!S`Q=t zHcKj8FSjWZ6{UB^e<=~JA~HOzPbkjdR|>ax`|}ThA><%m%n!TAniw1H8S1|pbDTHliG$MZCMjc@k-8INBTXFNx8B6e`XSS9!q+<5p~?dXR7 z*ZuYvJnkRxafBja~_gy<&Ci{)RC~DaBK>@U3 zxUCfvMFsn+c0y1yCB~6qVLE{whmV;0uYCW8N@8zVkOG^W*X6Zg10~>T9uO#bi39Qd zSo$>_lm5;q-e?nWP~E~Q8U02Lyff!r7Z1L@)7gRmGpLy4sp}y7EgUr8?;ly9K2QVm ze9nan{+uk=Gi-E$$eU{B=H@OhFM*>xAOj5=b$}o2@y|ZWc@%hX9jfMukpT$8*c%Y} zq1SR9q|JWwVJ8HKOYIpKEq~kL<#OKnoFsXi1j$NK*(^Ufda5M#>WXgDl~%u`qGYLSDWT#PdQb}q zi<&=oazvqqO59BK!YhN8UO;OU{3^JMY$>kON$WLEz*wm$_D$T{-AZ-?&X|OjYwlk2 zNXt!XS)azOw$YyO{GpEAbW=nlhV?;nGr|JL@>)x=t}nkpmY{U7<&nJ<$CmT^&$5?z0s0ZfG<_<-goTU*!q zy(6+dzb43A1x8~dUyK7IAA z?$!;A8b_T&>}2EMZ>wm?8K#x}N&mi1p@?m{O4hj}x<`(v2bmDNj&}K3)+K#zF>$fg zF1>5O$a9}Lzb#Mgyh`lf6u&Ehw~PN_S*Frb>3$ANO}m^1)kvc|bo{d9GeW+~`nO}g zi@yWeSX^ld77tuM$V4c-AFcBf-VmOP$*_=a{%s3Zydh&b4$|BvD)5~eC2u|IU~#|5 z?Yi(U)*wLzztd}c!^cPHM77L{h&X5VA+lp{Zx6NqKOJ;!5MQa{^R2UXHxzFPwug@X~24Pax;QlgDfB`zBvB3vX0iG%}G&mqE*wevJ ziAUJafRTgS?83oOnt-P*p9@HjTm9We?fjmgTKar>_9y4+f&ufCF}asjihE^w5~Qr& zf?oo7sN9OGE)aU!)`4v<;p#iugWR4Z%Zx5xc7x>1j)0CoO@d7^?-0O0_3?WJSo>?h zyn)n`W=Gj>2Dela_lZqmgNT}aVV1q6bXi_-)NqGUP3aENO7QT3{JLtnlMZm`gxVXN zV&x0-MvHEA+eD4<^fGMSBRVPZdzGA8f?}Bn^V^&v9tz0~AFE31Q|;x`Bn1;t({(Lv z2vaGZ_1M%6tnKUgWva$7qf)KAD((5XnVdb*E&K4R*f?xo0i3ho0@hl*O|4|ieo6n zwX+WqW0KRd63QyR%8tcoNnRaaFy%o5e#eEf+~BsmN>bZs?Z-tUKCVSiZuzC$!`b=y zqqY)=7H{Pli}_65{u?|Ja&sfNz`O>lvxZY%ByzB%EA(n=^mJe?HD|W;;;`JPZnfDj zAA-XBL&c*mBo+Y;# zn0{i_3VH0#tG!}A{dtXfjeNrfEo9T)Xm>*Rw8NV#fm3q9)!#ia23Wsa%gM>Py7GZw zG_Q9Guf6w@Z(P|(0fQtMpiN-E@_qL|yk(slsQ@T>1%v(t^E3*-Z} z$Q#H$8ThlE!z2$*n-_w!moI;k)$IEo%Nr+I6F=q@+=i}3c$(9E7bcoaRw2?~n{BC@ zOR*wfm7>cetui=HMcLQqJY4Ik*2`2z{1*0OR7qv%P5r62wDU+#>B+KuB?l!Tt&`K6 zV$Y|k(#4aOb5YZYLKr=(&Z660N$d4734$$(K?r_^@$D|Ve$dPn1g_+}4TZ!=cu8xe z)y<#%_r*gzIBYOiIEo-eq3k?o@*)iDpHY%D{uDVMzYS=)et%kp|6Ew-$}o%S{6<}_ zp|Bhh|Do2jCVO$T5Ybb*O&*x8xUGAUp&m%aAJl?mfGM8Qwy1j1YDu*oY>V_HnzDd^ zV%%PlJD$8BYQLT=AF!GXi?Fuk<2H)Z5&BnHYR}zOcYc>pC8v~as@B#+?Ly{FT&uXk z{@1vgvte&vzaiCW*F>7LP$AKR{@v&&v2h)7$44{0Woi1KQ^OnRVR z04U3*A?MD8b8XnFwI7Sci(MNczz&?D5_g+*sbBOfR-Ig>JKR3v7Qc`~QZhv#);r-9eVv)KQUI%GaR|L*R0Lf;WO z@G?xKve|ESqtV`k$_CMR&@(VhLE0VZn3(nr5TU8$aY4eJf1!mkz@e4rUaAhTqXJe{ z9Lm+QB5;QQV+RBtgaaYN=e~H(kvvp}0LLZap`idygv0N*yE)(l@JV7T-j=*$I}?Up zq@M`lIj`)biea->C8h^#^eF%yZ#t7hf0Gtz2Lzw~;<3qL;FI4DTrMLLe#3I1y0r3& zs=k(yWKUwH7rT&13B&DjpR^|+zoE8j2{Y@o#6XE6*!xgK@K`A)FOL3%4*rQ2x8Zb| zBEbZB0GsUUaVKe0Oj8w>ge=@FIa~7)bbXs>sR*j6HA7jqjZZZ+#WTS?pD41`-tE|QD z@UV%&=i1q3NCMq4o%COtrH>gQ?YjA-R73&Hq#$nxayZfb5pCzHOPyMLf=GCM8H7z) zJCnV%_WGAJgTJXtvWc@niaD*iGTd-op?l!^;I<7}OUN+kQr0K^GZDHtDV34MTs(y{ z)1f1?H#R~|_MC#9oHgyaM(%$7(PKavqM8&A@v2-)mNAdd4J-7wy{(-VO}6Dbzg$n> z>c88}{s!!#T2XLgS`C9)vs=&VS4)1MxGiV|UeW|F%NkBJDOPeW5%BN0$sO27)I|9= z*;lunv_-9wV~ELq9Dj|tF&T%0vcVNdE|2p2PIm^ubl&ptNPmZi8uY1vR4UaGT3ilT z<>M)@He7zjk?|rs-n;l&k#9=3-=;o*3e<9h;KXiksF)I8642a%E(J$JT^-<=pyWJ( zVE-|yW#O%!Ssk!D1ws!=S!{j$@rDVkg+{sx(~J2vvY01H2a>)t#xSwC^OTxS&ZdOw zOJF{JEkrn75a9;F)G-Q_VCYLpQ{Tp~wU_K>-45@BBj#dMzI(t3slMvE>DLMo1m;nZ zH#~!ONZ_nG<=(IDSl4lmT=6zqXF=MoYmPrqbYwDN45cu+kgR+*{$Xxl5+yv(Q8}4n zjunu{4(ZpJQzZY@TAlLhQr^Q^xA`6}eTsEZ#YO**cC-5>0<`&hk!xA;8}#(sWFf^b z04A-o?c?@^MhCfOb?}qdf)SEcvU=b?ij{u=LI4cfTPf)=15Bt5GDX2n~-P_f3A7GsmJKnQSoPK zjuFx2fPHgEFR0g%1kZJSYeBBM7!&~Td&s_&5K(?~Q|6~#AxQm#wAK|i$U_&?5H8RC zvjs^@_A5f)V82CX;(Lz$Bl3O(zno70l(IU3;_@`O$h0kdikgv@jje^S*rP>`m{_XQ zFHszV^HEy-g1PUhrXS`l8xRT-PKc~z)P8BIv6uy0T+ABMHnn_0%X~;>o1+vco{dx8 zjf|G$vsSgr*+y?RqtTJ+tgdUZ5B1 z-f0!D@e8{r1Qr}VE%NigYFmy2R#SGHU8&fGDa0w}@wgKe-2dA`{8j#Wm*9YUQuwz$ zrpxe$nD=Vyj&8SGm2AIgX$*(j$_uZ{nTh2*&Qy zLbXD8DQ8;~hlL~KTAcf%cYmU{ccj3bf0Z$$1E#8ndBu286_;Uk{E z@y&sZd~Q|I!*_gLy_QID=+k22rwgHwnca~IT?U_yOoRQks?jIA!cW<~B6?VzZ2#17 zs#j!khUVHc`OC#8IzO_hfxP{#hGK_pFd`1oEx{U3)CGA{R*3mCX>=3{22DpJIyxDG_lN)&zk%U>r#K#T2Xsq?h*OFv5z;$9^`s= zyzJm(2%69{+8}kFdMgjiCiP=r1~O@nnA9+FCRcIH8Qmni99&dHpjsn&Z#%8^NqK&X zA)Wy!0^KI6R0rkdxkZjf?TF#!P3^V^)!}nO6j2PfAB{hu?SLvJ$e&~Qk>9V%n+LeCoer*>v0H6|$W&2wmT1u@Qaw>I&%p3DhjG{ys0g}eb5ej?( z;CmPqCIHAVTqi_|B(BivBeiJO?%`HDR&X}6?=qzjR&S8jqvc@XJ1^5DX4W&UL16Z~ z1?7+TX}zsAvEJCGt)DiE!|~u|q7Vz(XoR%97wvbsk_@h0Q&Nu8l#L^d#z?5DBx1_? zCQidheyg~tPel@dhUA%Q4XYnmXgkRiXEX0f;qm(~2N-`Je+`N$i`lF0v7M;R+b5pC zW->d8M!&KxFw-vZpwR+C7PoN!sOwKVR@>P0n;9!xN_o641EJOnQTv?a)RZM_@6qq^ zP7Ah0Dkrdp{;7Tg9j#dv7t!H=ct+o=-l+13=A}ONVTlxfC$t$Bi*pv6KRlH@6f-ZY zL(r;44a$~2otBkZjHC4{{0GkOkO+|b6FY4C1tY@EY)<7&vCro(tYh3or}xR-%v?EY zng(O$l8@MeLQ$As?DZ7O)d^siQ8XVnShx&S@-D~niau0-Woj6~iON3K*5hgS?Rz>7 z!2z0$h{Un61URsathcu}aL5I$Usf6ElF}K^en%BBeJ^2Ekr0tl~_K^)M?)U1NQS>4J|wqxl8lG* z(}<*x29=p59G~8ZuYPV%n6DjH#Uhsjz2@l3^k?ce!Irxxs<^2YZe@rAheKKG;|k%3 z31&ou)*k~+9Y1f8$>Q(5#w@Cm{)+_+c=6u!3tEGk%_Q+c>%Pw)JoNaS*cfqU%@k9( zb8>sHBmZ4DE8SxJJc>Z|LL%o%2Q^z%Ua6c-7=b7{VaBoj=$J)mlrL3!4P7^ERgirl8vPCdP;oxS)34t4LZDF zOiH3b2$X-#lML_& zbAT1O5MRS6!7>Z^>e%sT7c>ReG-=FdYQ<#FvR>ym+Kj3jrTDVbp)#4?7z)7;ZOeL$ zH_6qUIj!kJx;f;NKi+Tm6c9Iq3f|6qRt07KGwsB*)3i7&C$g21b}UcC8T9aYDK+dPq6=4fU6CBg19@)ae}tqh-Q=Q8`1>^Z=-~X{<OCq?RsbkA;wP`k4H%RJ) z-qPih9w$UR9inA{S3m^IH)hvxkN>=F1y@qb!4l-JJ0WV*m=g0@C9<% z`F7tcz?1a_SPEohWO#UZ00wg>?04wk1EjWhQL#^?AU^y{CtY3L7MsQ6g~}$&pVYyV zYVd(fH#^S%$~EOZ}+GK^kN+z{iLY1I|^%sXpjW~nWgc4{~;+IH4Q`IRiXN=+mys!!ga8=Myq zYv@>(7qy4Z_AR)ek9Gq@rqrq>%{;NZ}Rd-HMW$;k}^fLWDJJvBMGH?S6X+If6k_`lADALv{#r1PmIeA@x| z?x)!GRFPavolzfl2Rit%7YP&>9T?&PI6pf5Mt(ve39QArX|9SQ9sr_$xkk?guo}KI z(fS_qt*D5e{>KzN_A>v(}B zkRTwsKKFk++g=V3q_W(*BEBEt)&DRcj5+jr_HZEfGJxpQkzUX~ZjnquD6ZBa)+fr3 zxjV;TH42OF)rCdh$dw)~epLcJ@BaJC!=ciXR_bXO9GexacyT@@I7n72I#;{yH0X1y z$!40YUXVO;-CB~Xsn=aA$5ljA%wpm9q7p4-!TG-_mWVg}Kgfa`B*K6`J6a@cP&4WM z)S{gdnT!>-YR7Dv>orS;&xxZqxAFN5!rMi>ugBc{-kZgRZB^d8jeY!wY4a*%VI=z; z+bd?!JK)1r^Lq*ti@whhWhR`lNL!bRoq+CiJDf(#+TVaQDi0mGEHx><@EH?@v&MK> zr>?9X*8PTP9^RnaB*8S;cDUfzhfdR zx7KF2w+eRi<$MsrHoEnPFLX>gRodK6irlb z0F4tAB&5BA1BO${{H^ly6+R4BfVb53vh%? zCFBUwuhnlsKYwd8_7)cE1TYqm;K$$-S9<+K&36e%JFC{rb>AmMNB?Kc98JWwb8mF1 zL=2gz&u6_+%>XbUQ{@)Bb@zjrT=0(oj`mN#RdY=o>PpJdwjkpW7!#5~B_u5bc8-#0BF4(+3D64B z7Vm^n^;jEx(|1U2T8w?h(!ShzJDnz*%p{M+bNc1W9D&!KSs6#&deKQ%1!4QU zH=DRH3g<|Eh_@K zfsPOMtJE%ml1R_HoTv_GAC-g;N5_%5ju016&TrL{j&%$}X%q;BeNpP1luP*VH#I^| zNEeY=oIF`IYuRtLbJ?nM4II2Rmx4uZ_q~5wThZ6|qNEsiNS-e@54D|`vNU@~5wCRqVJ`}Tz2AaJ*_Pzcm3JHi2dY;s_ zefHc9TUpW7->i73rT{^=mHy{NhE(d9AF*SXw6_Zh4Zh-mb!?Q;9N;MP`J7BAi8%lh zRzas<{|O1V+HaDN@$>M^Aq{C&2UPmue4@UifSq$MN`I_`g8|YIk)V&HPy!>%=`6qv z3f35*Zn!JDveW|Gy-Pr2~*`1iO#7U!S)D zyS>QmU;vfEC2;Oaka%Dg7A8a3$+XO?G-%`852N*5XfM*d2RMsN*&r6KwKSy#y!HaK zZ)g@X-nDVM0#zT3iQ$zj2AWbM9aDW99T!rz5l3p{`oTz6Vdc?9U}dGeQ4+lF(1gY%A;cHL>t4`^nN} zWOkxzyEH@)u>yca3h+?8zok-wK!KnSvQnB1#t5a>d%1(LQ5^Npk3@~)9f?}O;Xpkf zRW(p=>wtqIo*;2YUJ+IGI$r!aa&bv__`2+JxHLq#R;>Qf3FA6W#pwf(xc+u@9*L$g z9}fMC;pO;5-;E6pbL|)L3`*}QMVuwZevLp~ohQuv%#JdrRsYw&{KNXW%(uC}iyyEX z@xlb}zI|a2%xf|1dB67r5vTe~P^(RoU{3fG9qnSy!S$fGxD$(X-3;p4PF!QIc?0Xp zr4b3+P?(x0773a0iFu{x3P}nd_iHzL3QBJ9iQhX+KaQfc$;}uje~X_aihZs$KD}(9 z2lWdXnbIZ6jU+2r9W=A;V3_sBvptbK^zE#VxoJF)e|{UTlGEiiD%CX~b-QP9(_U(C zwkQ({U(m!KxKy9UzL>$k?D6HiF%IB-JngknywfAd{ZxPwjLCQJ+GJ+_jh@6=rFOqw zwRofTx*jenM9Dbzol#${Y7?i!g1v<6N+gR$8AwnPFqysUd9Htbs6Sq;X3$6Z2bTQw zKOG}1ol!-o zv_3E(O$*rx0e z^Z}GvF0XV%PEJ)?y*eZ z^|D=<)W>-H=q+<2)qxHl}Y9Vj4RzHS1^4S>@?I!H?4sHpszyQrdw z{H;#i(A#sP@4mmAk*Z1R*1G*(odVH%{N3vhH|zqN#?DYI!a^D96IQG9{tEkC19{rX zN2Bdr6)v#|ieyC;z~y&6V~)hcSt0pAfMW^w3uJ1)(He06H;%&ewlrTNBEyg8)ezv5 zfSV)JO_I6*z9gOmOU6B4K4d)ehd3@uC<#%3)xsYnAU@%ws>0;A5bQjhGCo>f;>u|V zwoedadlS%YBE<^jsDe7(ws+CN2H2E}STx9cC3#%$mfluol2jhL6r2rY5JcUSOHmW! zM1@lh`~9tNK)%fo0A|jb#~)5~wx&!|hD$YO+l(f&HXS-TDE^z@-L?T*ua4YlP9J^I zEp+D`VI)4Yli7RO#I3n@E3WgvY6d1}8yv2BykKeG24|+A1}VYjyo7|U{g}VgFV`c{ zjKB+g{b94+paI$#_GZBo7jn3l*Hm~7Zq!az{rOU;8Z-0kgO!OI-jfu8gpdgLiw)E7 zin^ml=!~^=wms=}`(}xyW?4v*IkV{X-}1Pf$qO zq?!Hx%tbr^1veSw3XtuwDJ3CA25%|Kv=lBSEchtJ+x7AX@VX8%c#!%!K6c``?GB!K zV~17lavfqHCTh9jp;B8!V1HkKsP#EuUk8W0e|Nhz3^AT4S?G>ycntI~xlTD%2lap1 zZ8+@8{qQyrwUE_9N)#t&!-vm9>B$5$4tleWoyW|)^pP|zR1rMh9uA0gshSP6O7i2# zAWZO`Ks_b^yKXcEP<$bB2D?60SW?*5Cy3*@s-iZFrnHC0i6n8c@zrq!fZitFmi?*1 zp5sM{)lkL+w*vZiZtabn8%eJ4pF_jh4av|XhlChOkha3O;YKYvKXVIm8$BWM`^ z;`ZFAiEKyU^C&t`^rutmG}G`VCJx*On!-K7qbN09oa7#VhH`lE<*yrXyu?+38nElm&+q5WTg&R-1lpu~aQK!kgZBQi@fd>QS40ZssLSfNzx zF;U=R=YK70fVGXDp&=zXxrp5w-~0Q+;)IVF`GHj~`%@p3mAidzIh)tIJRBmhsemzu zA5gXhgisa%Jrwyk^7gXLJjeH|18}d1TOh+ASwi@5__m?ZQ5;e(5Rue8z`Kq_7+?Zz zmV8K7%@w)1*zN=9lFj}oVk%}q8`R%DSH7D7<&z%73s}zs)VOUxfE%NS4;u_Go-JS( z-Fb%?ZVgzvQ1Ib+XgP=eL)0`i8yT#4-X!KlPFflL&-`$38(?2%cn|pP?vXl(`}{*>YnOPUa8ZP|eVuP`OMi(RRaw71-|$$dQV|8N zFc|b{#+i{5Z8q(2X;CM9Fv$~s;Joj7{0LN8rBP)_jbqe^uL6&f1-rhe;}oX;4}_UK zn$F7QohXn2M;B(ra?$$tC>J=vATbLLG>2^$t8Duapu-%s?8DOt`tIVZqX-K)uV(nw z1*5`x2(R@4HM^aP7Xb&Hi7}2dApJ9L{u5=wKSJG^%5d2sZJP+w^?(0WnoSZ0Qlo$% z_58J9m-p^P5=dt8?bzQm@|%y#fDS+=^yL8D^^h1&4LVa8F;aq^P=ms{ZfFo(BR{-9 zd}!B(AYHwq(kZ0s~P?q3ID8#>fIRR;Z1xdRO$x8V6CDr%(jVquN_TAO)4F3Jy*tfNYkYp+4? z=$a39(7p?Mmnikkf*J#dFAZSW`_!P|4E^R)F;ITU$PE0_DZeNDVbDj62?zL@_}pH^ z6jG^nBn0eiE)o$LTnmZcmYhhkW!K-haS+*j;XfArwh=sA2v}%Nr64klU(Xic@stoD zso`=OuvFK7=}j?Y{wKewJuP}ONQ(GCBjWG=yH)Ct)^YL%5){SN>xhw6oc{Kw0U8e6 zX?U&3Et#URu8Nw*hP+`;Qd%loa9O z@#rP`jYpc1`YQUaOr?xJ{LW0*7HohNCdQyX4FQI-S1z&8XFq}72&!rHoGE- zgHCe+5xApa5}-+-q@cinHw0N269+}R4*ZV9V*vFQ7Zrib3vmJ`09sN%Kzx(z0dm$r z7=~2BE`ny91i6so&V~PN7tqoF7P-XT{r{Sh1~Y|Nf;+x&bVIIp?{GG_n7IA91kdKn zak`1mQBg|8p_j&YeETuLVkp>SAhi&A1P2+x5M7MB74U{6A?~#{`1Q$nJFHru?Lru{ zz`cHpo2-SGtiktkMK91b681?Hhw+W2IpOD(-gj8caNI=Z=-?cUn6lqx?~c|STs5EQ z5HlFa7=Ne(js0eUTY}TP4DQ9JXR!VYu!8`^M5y_{nEZ(k!4-74ABmS5gbVpcGs$p> zAcQd_=5;=C!|jtED(68XbYA;lhu4P&b}{6e{j@_e{N8((J3_4o(QTH}OyK|=f{>eu zWN6A)Q?6Z*tVVcT461{hSOHB1=<*_(M>XrZ;ej4T?AHdQdWZK=36Hd|3a6zoi*JBl~1W4I^2g`@;}3bSI8>t$L>}~ zoo7G_roRxmuzrXK$RX|^fe(OtD;t4x(nbMfFj-HI=NpP+xOoH}lYwLdZ+ zr;j&QWjM*vk!L8NDo`M5R?4M6$X1*?%*?zIIa7LK+?9!Uffyw2J15ozUu%bY0!PXL zq0s2@&lcE3c8J_W3GlzkEd#*Z=s7^Opk}@5F|YweEvM`=lfA@Rfcmdh1QpTG?-`u~ zYknyeGrdC`*=#)^?T+N4cVqOXnU|j8k`Wv-ku`RW&ETWC2H#x+OMxku;Be*O!BM6% zW8~nyIjeA_L zm6MoD5c4@N4W~H{4&D~QF!+O_JIlV>!a~Pa9hDy2A&D@3~GkU&@5Gb zPHJeyBhR>yM};46zv;tAlDuU>^(f7){4OZmL{Ny-6=4neW#kX>dlCsWQ%qfhQes6(6An?EVVOEM zChVso8J|Dee^re)@rGhDiXIwcqG6@kC?ly zz1VU}zB^l*MDln!P^>`BzT}36$ zHCd{GQ-EtU1o_+fqr=(ai@sM0p+GgDe+fK=g8<*G@PGT333MQDR905^+P#?nO_IFi6YN%cWX849Kgv%B%66r^;(i4hhCMhMR(S8QF3jsyK0+HZ{c z%%nh-CKX}YWtN@^rU}S$wa4PsH74W8ED>1H(bkA)9j)bas9nYcO>|oTV^>E`8t( zwk8eY@Pb;(2kKsTB{?_e;$u3tGKJ&@7UY<}pvTgFTeC+-fRC^vVv;`} z19OO7Ko>7uFhr%v-XtoEL<-9}7)<^b7XFycAN|eh$(YKNkjEe|rReE8mIaowLR6|2 z>PdacVS0|GR)|?*dQPRC-t+87?-tX-*p0g=o<&OhNg319o{G=uq?Nt00)L<&*9%7xm49k(0dO3G{)W99CFtJl*S{G6-hjs+G<+3l8+OsJaT&YFa1HgJK<MjKS$#jlglzsu|#5JjOM-$rpUJ~a`cs8WFr)g^dpYkWw+Jk?9M-kawE z49qp9SK$k$H^Vmfr}Fndce4OoE=wx+s)v^zjOxe}T)J zVDvai&CgTa z!FY6pNIbd^AT2y_|5FxI$9cMu5{?cN5ky;$I=oFJq!B6l+J+ihNYq7HOUU_AKfgPM z+oO%}-_`VO{Dp?c%Q=$p=3Dr4pu?f+W5q8uNZD@}&OGj_0Ryno^^&hBCNa(PpuYY? z9P~Z?SeTs8DJp_O*{y9RO^JqnnLV0}iHpdf>umQbX2($C(JPNc`45j@+1eC64rV$9 zDka#Gczv1tE#w)18uNQw)yGYNlvMJd;S4&Ku&~VS z57V@>Q@YfU1W)!H$a+BseUJ+z^v2S_TrQchv!8>(aA%%(i3Y|Ss^A#V5}&+r)Mi$y(pHR!f~f+4tOWrlj?gWKi%KMLeDZjYUv&8mHa_F9e6j^GHi_ zrsat9!l$j>CtKcq-k+*)LgDxi7;~!TE4A@!RJQ{TWKpRoqp(Wrf2?8OscQG}VnKqY zH>0USVHM~%A3ZWDJ;QmoNBZYNA@e(w9R|Unf!B|o1oJ_6zCn=x;b@+lQGh5lTtNsC zM39FBtrVIBY1{X?a=|}g=AEcWty!ni4AWHa~EgU zTa7ad;p`x0nI7k3`C^-osBwg7H7lWpKIn3yAI+ zp?rYt%23Va}*XU@zahLY9lQDB&*w zFK+>~AzS*?;zw1OHxf#E;D>Y$)VPeajA;@)uVtQb;^!tD11v@b?BpHLELBsOTc8zX+>-Cwcif>185%f3JcW?UaBd4f>`qPbrk z^{=_-n((QXuS*n(LG%_CX@!BR-x47vgB3)mAeI0{)4x6S$O9T68XO~XGd{6(ltg^vI!l%F`NCAPpnL2I?9 zE9CQ*{wrma6g^+zmFM4^KYKNuQ)J)2F4rzk)l7$cv3AoI{QQPLB}G~LL;S@&nYW?@ zpT{~8%?+=K9<{Yadl2qZVL@S8X^YC$lXo=*-v88NjB&l8({EVb*ZvkH76HG%hV8HH zSg9@e8Rk5L6QjJhJJan@JCIMuWEa}*_1+h^e}jrt`>dv&E&t+=>ieSj^*YG6%s()q zb~oUc>`ybl-SZHDbOJUbevf0{$`)WjF`|v^-z5c0EtCl7HEb z0Ol0D1P2CijZ~0UrGGE#{6U`lHg_l;_M{6w?TxtWlp7dp*Isk6KlR)bq!@{vieJ2> ziu|Lb`YjL=R7(GuBWOD!6*hVN0sBk(yHU)foCYrb*9<7rvg%*Vo^=IDXS7E`&Lhd+ zeMzs>nh`LVZrx|@3`8~a-_C4t%fel)qYJGP@@`r!4Xu2XEzJf}+JwfH&% zub`*upw12f%;GaEIECmlilL7$HGZu~=SRxQ(3_~^QjK#GYwg=`S1rK_Q` zVG$6hzQ@!Id$_ye5rfA_AWhLhti}`K1h>o)b4yQ7K1P!7{43q`C@6$a0O1{DOsnrKJqms11~p=;|l$fQ@4T*?=_6&(C+AWV$G;sGw+% z=Sp9JZrJo37*K5fzp(>m8fuZJ$=eeQ;drKaM%O%a#3RqOAfo+LG0*k&^#PtGB4z!% z(+5FRAw@wpLAyTX7K?E?o^A&~jaejWt2v6}iYo&^0?uj(KHcc4$J^zaeAzuQ;g{b) zCR?HbGbrfD?%V6V4s0Ccjw}*_bbug;6~KhH{VD*DrqgGN09HrR;BftPdTbL|B$u>e z35mooz`T_9{(D|`V89erl*JzB3WjTC@`Wt@?rqqW(Kaj}X(4gonw<3Fzqa*+D7qqk z+uS1t0wU+=A5F&xWZ2%Sy)AAJA?LLc2@&zK9LHFAxR`~ z5J>wDh0BZ5BHnuP9F&By(;~Ei0`iYU8uYyz^zl#vpT)ky95*8tb!18Yo+37pr9!E< zWvXpaOLH6)__9(;Kv95hud1#Gk^>2h$Oi&K_W?SD!28dWOKJ!`g5MkZ3UvdG1eADa z@lgGW8Oiym#R8eV&w#i9_y+!avLx4K>R+9py{X&PXmt?hdcU(oM<&7%*)IG}6y$5` zDBcFT>Tz-terELrZAI-_LioZin%T@*+HQl{OYUwf7zp4mw0{ql>v*jn^u1(!AN@C; z8EKTp;8FQZ^=kR8#l_eOjW~g&w9UaZ{jejT9x?wn*-wn6HO3@f2Yvf z?DeFai;c$pXSnE!e+UY}dwNXQyIB7dc=@}bRV>HPlAbup2BX2=qcNIF&NJXcZ#Xim zZ7Z_3tmsM@ho2YVS33f?JeG=B$77)rz(S z7%AZ#G5Nbqp3F-S#Xchqm{yz`SxmV}&9oXSl$wxE!W+C#gn5|kSzy@TBeF0E89tsiGdPe&N7{TQ>zxvFn{7YorvwzFb ze2h!Qzp7o*DP<-IX^ey?#gKX_!Dhr?jw%a625C}fs>g2<}_Im&B8Ij0Q0H z9rACT6Sh5|F!hgwGfd-q)Dy=#pb$BPoxv3Mh7z!kG1u@1F6 zO5y}A@R~el>jaX8kjEo^wSTB>oe+VxZg@CU@?ek}QTg(QzyDCjRCCcm3jqtM`E6BZJJ2)I)l$U}Ah5cKP$j zqV~(_%ii+9NyvV&C+={v0Cgz=gq;Ws^N45QX&|(WKgF|}6-^0$9#O(fBf68#aTx3% ztlHE?7gcRFUT_#pfUX(CS{C^D=}?@;*;8HvNGlvgG>m$Uo9cEgDLv&Ba+A5n1?dp? zIDX=f6HD`>y}qWcCwqbsZ9hC~S`z8>gR=Tov9QRx|9?Iq?Mzm;F;i)`3_qF`?Plldm~&V>91#aY}~&im`lP7Ro4k}D* zT2bQp?+eGN2nYx~!S6WmZeYg7CVJ{5(O(TAcg4SLYzIJ=+V&gh2j0j(oXJPzMSKkZ zr;i+fG)kX97%>^y`;3e$o&RXlj3J`*4rpu^$YuUBd(buj>2#e# zmIOM^CoQ}>$mgJUrT>%L08z6975mvIHQ$%svhrPB2L8E*ggV@hiZoE zCDqewkE+tr!SCO@siH7JqKiW0=gGGzf_#0RZ#?22YN9IYEA@QXwY=;4Pk}jOOV&FL zOcr7%lxGoY#J50uT`1Mjqi;ELAW6idlBJ)(8l>GBt%hAOD3myH;gzrEB~ZOqR9^a6 zAHu}fRdZJs2pNC`V$Fc`5Lw^jedLfLB>TSTQ^#^Gy?$HwQc%G==okEilu^?L8TUee z>>AQxp@8~C3D2NGw+3^F!8xrYxs++eygM*yk*|K1V$ZSPWJ$(hP@(!5dwKp8^J<7L zT5?~931U=|C3}2^Id%#LWmn}d5n;ra<^}8>{6)qQDD-4R50oQxVVjYWQO=4W&jA@j z3hWTT%^NVKLUI%jZ{S!BVvn39MsT$Xf7~Qyq9%)aC{XDO!vJVM_rD-o-$V~01Rt_z zdAd`oZMdTo0nYb~dJpSYH8bdRGU>h?123o-Kyrt%T9**jKwy}Q4rm5|xnN#u^Gk=u ze66@#~)2I(=s5%sc}dX#8~ zz8kZV4z8A)BrGSK($-KLnk}|9438^@(TFqmx67D10rVqXPcW#r{M!hh^|TFkw215c z_eN+pQZ1sj-;(BwWkWP88y!Y3+UdM&!-Yd-R>`gr-Lr$_Z6rjX;fR6K=ulq3zwe!L zYu6z9c{qT+Q+?fg;6Kx3I)f_COO;$RC;dLHAe1XbMPQEtJWeqRVl89VDiy>K z?r7dk+bSq1`1tr}XkfR7fuc`l-QWGM_dJNR)w|`@yV2FZqXOZ7v_`~)Sc1_8J-XMxliiGM=E>({F4i`XP`4mK$A;@FQHOwUsaJ{?b zLOw-UqG+vU5upqKyJY~R>%mhw^2PovzG@rrdFaXjVd+}aD+`hea{p-n+UWq7mt~DJ zPrDf<>8kAN6-2caua!IQZx1}n7Im2f^6d!A&D9ZMES<cgEGmq0}@`~;wmCd zB-?G0*S4UV^*u&w$f1Ck$L>w-#l7}}=RKvrssH^5%NVi%x$W1rL4BgHAI8I8;n!c- z0NyfU?#o~NP5rM$sR9@Ut5@U7f^KX6l&qjhdbtO%7SsN{W>)lno2!TyL@5Bl*_6BR zN!YfoxZTLX2e}U{B=o&4NL1RvWNs64X@tW94nO7IhhKI=Wri0Q9s=)QTqHK?1$!;j zHywH?acHQkV`8mR|}Gt_Q*ySk(`=$R2|iFRfxwzx~tC@(U(2n zVV;p|zCZfPgK8e>Z*asal<@Z%k*FwVJ$P`#nO_gg5Ey2O8*$!;jb-I@CMtcdP;F;(1z=OusH2;RsVs z8sTED=JQd@_RHv)Dk0`2+UGbb_8TI|T$+YmF~4p)XKlXveLlK9I_HJ%0`WZ}oasT0 z=}2u`Id7#`VZFT?TG9L{FUuRzc4jA%_@-Iv*0}ZDn70B?`86g~etmNV_{lNuPMl-S z?W{NxtpxHoM(kLq?N)r0EC+t6~izdKe2pD8~oKKG{hyz2rP!uXbKh^KE|9Z-?l z?^ot#kc!6A7mr{6JW`5#(mOhIzC%YElROUPfFcPIcu0~Z zOsi6bHl4UFfc*^8!-&+k8R~)r?M@a0!j(5^CV4vm`Qg2%hP73}iq_N^!>}~TP@cK6 zzJA4`kAC`?QlD5-p=A1sEji_g9?sFGq^0izavp2e>+$Orrj<`75DL)NXPE0C+I@L; z7RA@x6LO}{9zNCVw`<_&$o4j_H)AGIvogh2h@BnX7L+bt6JaD%!tX&{6}BSCUzG?Y zRU58-d9pE4q$Xfq&&khE(0}&@ud7YE(Mz75F?>5I{FTl(4^Th?p?1hBIJxZX>@uNm z8Rf%(1e}?MA{&K~WMF6rMMXv8+s?L~vK@T>ENp8(t#})*FLn++{`u>dm;#3Oh`ocu z$@_3@%w|mAHHx=1bDT0VGJ=AFJUqDcbDmqPWa|GlKyKCvRw?2A@o;VrH2p4)$=$1D-B#2;vMPJ8Gqoq8on6j6Zk5uQ>q zYJ+Wr32(baV#8BNhtSb|?!3H|WsYA|wr?cTXI@K6w7~e|Ju^jvv$ng9E1F(G+$XsG zJKM)R7V}EYtoLOoD=iso<^R6a99&ObRAh;;MCY6sdmFCHjyc3&&Z}dVH^Xk{x%t`% z(1pV?Fc}fmPPC2U?h_YHod&! zX{jqI{8*V&8Fhy(@UJfj-bOg=Fv(Ba6IF{A`(L)SoDJN-G_mq+kDm z3@bW}y}hukchrGMu?=pdoDS&PSY9=l=ygtSUYIODYj5D#bCeo9=Bz@H$TBu!z(4(A zz;)5m!uCC`U7q(_EmC`vc~er1lS5lz{e(?Q4th-dv+wUnH*H*261P)b;U4@Ce;Hrj z2g9nUvd^kM#X_UWMSSu}9Mm)S{+N|y`KP&7d>~2cRWoiw-U?5|(uU=%yIohF3=kwz;icaY*e>cr@$uZP39jo(!hhpIu|7R8& zzl`(XYJyz##-M>XiT>(OBXl>FVl!N?>sgv$I=I=d*z1>C-dVxt>_nPKXnJd=-p| zIQq^FM~|2UX7h~A-wn!t3ygso1Go86W3viojr_lx|3Q%dv={3@9_CTeC9{yItd)1N@gMiR%&ZLA}@ z%3?Hwx}(MvVgp%VfGnML?db_0A3dL@eCI%jx7RtXFcIZa7vac+XJ!4IoxMrM?4Bjx zDk#Ca*(fWrj zS=i&_krwYq`!Y*UT4fZnh@!M$jPS1&LoJkO7F?^X|VW!Q%tv$)F$nCEW5<~QLB!?7d6bJ`yKPY zcQbOtJXT+x-@o`8v)a^$y5hp)`k_t(jEXEHVkuYkYmJ4>2nMbFqHDIr0&IH+*u3;k zEAjI(=iJ(w5Hr3TBaBax9>I8l)e6JUY&<5cIUR4bj*?Qq_a6!hTyE%Q z(Qlmf>2WWzJ{x=5uqu2O`tJ3cmuaTr`I5C%iIGcLeSD9^ZF7HhO49=G>bA9}>GvaN zwVijok`+UIZTy82D%R}1LgA;WBOa5d6@NAdZT^()ux-qK>={T6h4v8i=8->i@&-w| z(iKn3qqVTv02MjFsy*;FN*x63YNtrH~CU!?oW@k6Wi zYi3IIu*ZRFH{40L{!yCcaeW`S_?mR&p4@%d+j1ix;~+3#ip!`G@%hl|V$_a< zAKlmZIk$?9ibxsDES%*FPs2Gv(&C{v-I!bomwhZ~xwhhbgSSLW+AM&Pj}lxM8r2Q1XxGWkaj(3vix=Z!A;m5w z?+hv231vUKY_CgdJ$lvuftUyBs@7H)U1gAxFVf<|FN?1qns(;0U-h9!UlhGeet3}x zw^!Fvj2|swwxrmiieg(^$;Eh*%%%bUL&}k7VK!F@AP92f-Zd!2>&|VDrV^4S$=^tScOk=h* zM3U0~VZJ<>EY`w92)%1OG?Qa zmfE&_nPj_03#`DMcy@9U!7AssLlh#UoPcCDz~FopAkiBV1tF(@hY|`QWt*eN>*(kJ ziN%|#{QGk?`J0KF@H@8=YC%bqm>6yZr+Ly>!m4Wg7j7u$MZ^pwkWfs>!nH$en-he2 z3WIIlMn|DX!2#<`oxEP*_4W6p_xTjovkSO($N87DQff}KY)O%DAM~5!lg$Z3%Fd-` zrKl|{8Zc~i8GU^|K)LFCd%(j7iQ8|VyKceY%MUF6Gj)+8d*@@=`)p5Kss5|nt)_2C zciY2FnQ7RCV`%TH%+y-3_~Hv%!GPY)KhnKibf%}j-i=I4Y1zK8EG$U+<`~mox*)3m zWzso^wls@#%ZD6EWU)Y$Rlle4WYm5bLgNePI!iBNhJ5n56`&rg-rsjzbN0^DT7Aws zq`Zm#IHOTQMdc~8B)7_+S7T=G=_PC3ZIZ=50vO*wvC2M9VsuNQ7p}#TQy6NJlGJ7~%+SlutR;GIV zwv9sceDMAFw3jEI2eqkOio@)8O2<|NZ{eho^3=XIx0dX{o2@btIkTDx0fvVeha}m{ z_!`3hTPp%(fIG&A6UK+thuiRnW5S2yFZp+PSmdQ-zyHRfBEfNO2f?g4Z^1kF4Fp!r zKT|ge)C?gWC^#th{fYFrgSTIOyv{-57Ys|Y)mGBSe*jLj#Ps{|0&c$Ha8$pyd9pr4l0vi&Fx;9_{vKM&>4FD`Zdym58mljP)-4pM_p$|MIKzv z{Sh*$rJzas>1dZgfE(`bXUr{@P_MKylKN@wN-Sy45PLD_KGzb}Pt zf8*9Pp?G2hg*BKxQ=F_3JGwECug13s2<|X`(JOB5AO6KZN!;Sfefrb0pN$sASRhMD@+YS?d(+_?eX3P$_=Sdl0vIH~ZtDPskf*&!Ezhm8MPbQ( z{RS0r1tH@6H*;^^X#aE=d-kN?f8Ll=_->Mc9CK5Xr&3#8U`U^m!*?c~Z}VB*QmT&V zQ?vI6_I!~f&vF|(%RfDDb9K%hGj%ao>LO@?uZtkGy&sS(}9s>E_~BWuBcE^U4AnQ;}qex^=vF-r0iPw6BCC*>(c(Z&zpsP zkK-?}|BqUwJ8DvNGAEvSmGBO&BbxQ8n~uss{CK}d&~*4k15gmZo_^=4`F^a~pkHI% z0m#-L2<#wlG4e3j4#BFhR{#C5ClDR9^V3=2VTilW2|gR6v;76&?JGS!Ma|8b5YglM zj0{Te|0H>(epCWfji1;uh;)HRgBpZJG)LSExS7xYCBDi2`!49Wg9Obuly?B?PDw#g z_-1P~CrZ8?+W3!z{M%~ruq&vv-KfR&q8v;;;x3aPqwz+aoIfcUK7&vnn!B*|+ES4?j?hL%D&3q6$5nm&ix=p&-J~ix_23dP!l2Q5g8BaMcQMDu?14gC{*P z79la5;{^yVpPV$FnN+_Et^F0V%avg~9DDQ5%F@@QI{S`a3A5V#xi9NfZw%DX9G)Vv z`M#4hOvVX}(Q(17Kd;pOgD^+;vBmUpFi7~MZ+v6jCL*NQkqY76O@7NsLaZiq0gmxp zyMd?t*0}%O$ot!1&Y^zHYjM^AZ}DK(jcwj}B3?+nOn;&gGdk<$vvV$g&KdjnWzk02 zC7<7yWjjokzgPAm_~qL}&a-{N_)+@>ja3;*W|2@SiAKC#kzKJ}kG{Nc@MnKBit#~> z^H!YPGHR0WDSkt)l3&@y5|~nKe>Zo^a8IZMh50z-L^$N4>_s=Yh>LWKf4-w<)&1DL zbl=AM-^eM3{_Us;e%{LhV(&>yFPlxTMN*Yu#IA}qGek^hajf|0ce;hwjHg#hsvnE| z%spDgkS286v2BK%S6<$JaF7E&Jk2Ur^V%3@~R3JYweA z^7~YKusbq0WmYFlKgyx1xEWu4l85wbZ2HYwz1up=M7EqSL0qOmp%Dz7ycKL;XUh4Yxw&c^N zf52|c)$YgaS7^YI14k_VA6B(QWz^)WK1v%uC^9*HJmUZM9J(Qh6KqK6ukYj-NFb8A z%~qm(KUi!!0#&KdkdX9YxVDr#3O|PEzlA8}7~_+@Aa$kiS3i9hb$tXaxK(!tiS)U_A<% zMi3k+rMf|*%grHoSg&F+U^XT@YJqZ4=gpD?he5SHaO31HXtsFp{fDg);m+sD%>}gf+kpM)=~Af zxIFs1Q&!A0ijo%`0j~!`xb3frmh3pj@OM1%cDda0Oi%cRW?1=BX6R;MQ`(eXsbVRa zTcwdp6vk~VKLfB``gSp&hMFc1(T!#)lhb+^xITZUd0nQlL6*27^V~W|h{YlqkDFmv z`+44T_l+GF;P$Q*>vr_R+@l7@aOP-3IlhQ?)cS~bD3>PI740U^!h4%J!$gHJ^`#F6 z&CcI^T*BR=l3o@m?!tiib{b?nlHJx%{if`$DH&qF_m>>!Tm02bjQ(7p=^YmfOK|YAcgrcQFM@Q`$$R?>#ea%%U_OiwmN!&SK1T8;+0) z5L;W+2{min%IUQqZ&m0>SIvL^ud3F-Oh=WbmgW18Yfk2X1Vi&gCkN+!4|&ZB{(}C7 zs66>Y_1+O>?eve5WDln~)$vCw_iF_3t%vWpGY63TgoiVDik{3#uU^<`DmZh;vc+(b zf52-GRA%l$aN{2WGXv{Ho^j28IGqpFccCNb8(;l)Nsev6Qx5}=O5ez+7ea$BFr+lq zMRW=(IW1roaHd~y_F9KYF)=}?h~rx7>(9ZtT1u9n!sqv~$Esxw{~F7$6JTtN&~gzh zVOMu6=#xDevs$l`567jN=YLzoj{kBQ0Noz*IA0X!(=$aGeB> z@7UN_CFbzXt0SFR44nTxY4PFd77}=tw&Nwse8CF2vz`{=4J7?aB8<2=5`wSF)D=;F ztx?-@p-gLLYR=(WnLJ0wY52G$bP zg0#$)Ydq-B@tW^~Xuc7CEO$~JscGPT2G)#6PbK@)ifnh|Z^7Z@b%uxU9VRO39jsDq zanzR6_k?%eoSW)syL%+Ou2I&ObpPb~#sGD=2DO@+PSug-elLI;RilSJL=%DkJZhbD!4l4XD9-Ru6-e&g_thnm-%hngF6O_et)bQGjf-} zmpO78aTFdDS7f>b3ykrtx@^CWq+ZGheR;bkvEMtC?%#Lmm$K)tNPHrY?{--LJ>x;M2>?Q5rEtZsbDBr%;8IWVm3vEzYT{k3P)QTZ_oBkf6 ziE*Zpwx?Dekmg-)ij>$&h!nQTd&tc+PB%Gz#OrC1Fr!Xy?6em!^YZO7cr3I7Edk)S zE-3+Q!z=vOVr}B*KoY6t{hnMNH1tXU)%g|Jx&rME*g>N3W0(H^nIw+^$JmPMYDZme z@qM=T6V%QS(|~xGl9MAHc)A&OcY_kY#*^QT0!OWU`}Fenaxf^ZNuHAT2>dd8>1V(2 z@Lg$i$4K~^rZv#>d}jjG?O$&4UtQ6w6WXkNaX3&P%6tkN%H);3zy1lr5~zLzI5;#y z_GlcB!AmUZU$1}n*Ud##R8&Ff0{PA@MY$`=IV%x^FociwYr(kZ(52G%Dk(%}A=?ql zWV({^Ifde6AAX0GrmQiYayGrp!+pil$Li&$-ADP4xuA$$7V;Jp_+N3{o<5x^M>+WW zc7ur~(BUxL=38aVWg1gYH&wyvW&P2J($W~qKulSd>2TmKu03(>zF5Y??ycp27MK(J zINR9dv6S26fQi}1e|oc;chzt@UgwWRbBl_{VrcVyX1Ns)x|^oT_$5R>L>J3E=qh#` zm}J6E97Kd%@#5W8r+qRMu0nb+-#XWU2PH7_kxBBG{aHaFui68S5KMeoz= zoDp`RLzcPXn$u6dhy?Yt4OFt}W-+$Upw$g3akA-|!N)r-byJjAG{VofBOKOvGhyW^ zZad632?p#Ulhi61hG(nkl&|9(4l7Ae<{QX+9hpq6pW#E6pUrJ{bxd#tlJ7Si9*NtR zf5@@=#3T3^%>^c_zf~cU($G4Gvm{hc-i&>2Jn80?65~(EHJRfb2m2+Bk&AdYawu|- zCqyX48pFoqs~=oJZDe^LNxgfTAWd1bfJ3$Xc^KYW(~#Zp(nB$*?nzV)f1?=6lwL7a zz>l;)!hDO?Rr@I~p9yQ~3v3x_{XqY;X{v8Jf$ksdWTxp`v;+~o6M4{Nc!>vD{u&+* z8KsvFux$2@7;g*phKSy8gH~fO*yr+3y<)|pBPaj0yZfgbR{%*t09gfeu&#XoX((Wh zh=@401fd(RrRKV(CUJzwtEMJzI71c?8(f!LFF>?EkxnutFE1}SnTUqw&3HVe0Q>=T zE+N+c5d5g9kg3N|iBz%P$#=0J6T%x)FV9+2@0|RVK}w>>jVpoStj^*}x!u zPMWjwzEC`N#H?^U;uC3xn7C);eR=e=Xo;@|#3^qe7225z6p@n@GknaHb90;+m9>9F z$-G_QW4%TKtX4s2*^|SJ?{|zHPu`0gagLR<7{*UoUlQKDubJkj=H__Q!t3Tb@p%{B zU(;=V6Q_Quo7n^*qJ<4h6=eBt7P0A-&4p4u*NN8^%2EaoQ$o5GZV-QhwGs4njjWU2RKKP>X!XkKqBYer=`N+?F`tr4kV9%B0?}n+Tk2A~t98;H6%9#B{RyGx%SCf`K9BbQCe}KrjRTXz3!C6bIz+Q8uT)8G zJRz&G)GW3$57z|!FiG~ECA@_AIajBG9nW3l#Ol~&!n}EhaDlm|`HTsYV!hhlDP;(6 z&nfhx3N4)td=d2}kD>NDnljXS)NtID$t`YaPe*Q(goZsH{!A``f^(I6n^x?1WX z@L)ykM_sVD(rWcdOOR*lQx42g^(8}$In_L)XSZYO7H@E;-98ZZ=;SlhtP~Sa+Yo&p zCdH~@Q{n#C?zZ|c_S>%eHhCg=bxKpy_00|F-^wYhbX}z3d22{YN)Eb%x{9q|TfiZp z1;{ufnYx_U%wM7r0YNPY?D}L?ef#okR2+Qf76EWh$f4iS(J_QyV5>RoPKtU({{N;-mcS{P-u`Tyq11(_cRWF*v2vWx^jYAFYHo!* zk^c#yE0R80 zJbfy7mLWM3E8fkmR4b#3hd!ECcRfYMX0pWY$ot8o;e$iwSGD1x8xBP`+}LGnO=p4+ zHGAw##W`l28TAFO<`FB5Rn{zQZOIVM6e}O{z`Anl4QP9n0RfYm6)-iAhfEh=bC+o zvOP}9RNnnJiOB!Y#&t6M<930AMt$oMl>~);fWxp3ld!*hVVEN&-5#u;bIfige5mEtxtS9k}4C5AkD)S1-U7}O{V0xytZ5!kB0Ape=8UmOo9!T zos$9_g=9X8ii+S+zC2pZ&&mqi9;0`L%6tn1Hvu^2*P!oTZ(P+M(y1>5;~=<{&BET#HUPFh z9!z!Fit(VZ78uqbXq76Xj_AhaiO?#=YD)Zu@?2lCaZ-K8oxl(sw}w>Nf$sw?HRrnW z@h6I0t|{(S4MbJV=ESlR>wk3xg~}hgg#yQfAFsj{e59W7PvZ{pC@{4YxjUb%;^6du<+FVV4yG9YOc~#q0kKn?j{NX6u zyBNK!Mq8()mJt1bUGbo+0twsSB3Z>6EtK%gVEnw=Zh4Z~dcZ^cMK1F6df9AI3 zhCx3?pj89`4*#!8)1x z&=ncXbjXShYwcir`$zM02V3Xx@Nh3LFR7hcP^ScMynn$zz>ra63_Z6xAegaR`RqJV z2<{19o6h$3_P;g!h3mVcwBp}2wN^Ugt@XmC0#pust%~HKXOAkUd8$P)=d9ulcLh@2y4FA(S z?3s&}UQaw!UcM{d;N2OTk9RoOFi5X_zS?cq&N|%lCw+6%@0TflWtl%sHZCM|jcVDh z`U+-Y;Id^YP$w*`FX9yTE&9bz!}zDcRYgVKvZ~sLfUVG$r?l!m^~E8dcxAt7%9ZEn zgm06c;MJ)=IfU+U$>gee9l8!iqanBdU`@sBO^M!cHqLL(I(Cal%U2yw!opnRhFndW z?C^KEpJ2>>DcCk&aGL=LsUPoHtef?go}x6TW`Bg&O#pnvP!daS~xhv zF%oit;k|N+WCad8t%%3tu9cn5_IQ{(=cL9#?Ix zGt=)!F+NJc8l%s(a=}su{*zU3Ft62pfqahWdtA1n<0$r~;p5COO>HdRZgOQ!a{V&T zU`Tl|)#xJK0e@Ae>yk$3q`K3Uc27R6F-6{{8Wt~$tZ2&96;M6uW3!w5;56Q`aGJ)E zD|i(B=p}=?d)g~^PuE);T0eK4uvbED;=-N&O$uE|tyv7mX8s(56nT-4VC_ABnG5cK z4S~%a403ZM17c!a@GS`CTXWD*Nz2Kxc=ANTuKSgg{~jLgBfQd9fC)HW0BzZa+5P5e zBEN!M=@+Q+593E+KLBi3DV88~*#*AT22WB_QYz8S)$Dz5V`fH8PLBQ<%r z85wb2F+j9-%Z>e=g#~2==>Sym^^I|^Idjchj0lO&66-F+wk4~iTcN?CI z=^FekYmRVB5&1tvomE&BP`j-c4N}r62-4l%-QC^Y-60{VbV_$4-5?<)uxKO&X$2{z z;Y|O3@3Zf`QXW}r&iRdRjCVv}R!h{3XgOc0I%EdWl)A_UFhIP{wL4*%|BA&wGFuVg z%gg4N*~RDDR5RedPBrAC3M;q#Wrtu;x!~}ok+n!dmiGmdaIe2y+4Oxd4KkQb z=KR+CP6g=L0-+4QNM4x-YgKg)`@3k^!xQp z))juDM=R>Cj9Gt|$H_bKpcpaDcDpKGq0L~w-rb@jV(Y**rK+7);f4%3VKkuxOCRx2 z8Q}k{)+|W&KISH4a|m#fz_(kd_p&Gx-P7G@U0-$iL+3)H%MdU!;v2s4Mz#D88ONth zk`uG;;jbD2p<1}RT}*uSR1F>4KgbUINr85Ni4BfiB=AB})G`e^N2v1w%eGU1XNCz7 z^w^aHn7@$6pQb~1klL-DnUx}ide6+v%$aP?3ZR-8cl!W0^a;Salhmg znN<=HL3kfB6paoTen{^dDl5fk$Y9=yVV7?B{DIhnk*zJ~b|BB-ag6@t4e{TY3Yg&@}zr%=-h5x!(Z-?TC1#X+|@D#stMu)ZtdkIBE>UieMjo zqvD5x{TDOh4h=devolb9zO2F+R;s_8KXrgBJkboq>n`ub%URWm4{ovwSX4n-q7S;%Htgxix4EJ`D z$H{fx`3&@Fhh0LjAV0$J)OMvbLhP=H+)!7^NxT_^`F|IO%|;5ftQxD>mFd!@wW4c&&fey!D{;V64KuwJvV1~`u3Lkkyn77gcLFn+cTJj7B;)Nw^A!UJx(EJSp*H4HvL))re@<%>Fp07e zGLnT) z8pwmN*kANp9}TD`c6P7o$9aa+@LIH==bmQ_JJsbIp+10^iI*18=cexE<_DFB28GpT$? z{l>mN6)h&qBlfma!youUd)oRGO-1RD)lcQpw-?E`47!D=RAKq3kQ4TRAAjNW13+`_ z{HNy5)Pp;)k6sB3RtwlJYk50S82kz1?A(P3Pzz$0Dr6go3C5q;Z=9C=-SirRG%w$i z0W(7jIM8yue=UcW(z@}e8AJQB#4T+KUnstn80t*_iUaeD+YiHQ&1?u!w)g)nfr7}Vh#T?VC6{tf6-=SpHQNV?< zRAeImI7}4gu4=ra5aEn(C`A3|xmqmWxUlPV1RaT3v^%#>uwh?XDz2gOhLV)36?j80 zDm09MJarhDmHdSo4^jqm>I>fS_3f**)B(&QrH->$GU<{9R1ygt_j8(zLn$v3)0S~oc5#PZ(mZT*lV&p#lq9NI$z+KMe zpl(l7xT6prV76+*?cK?no|~VARNTu|Q}l(QS>R-{b%F0t%hRbTzFp&T%cLX7vzK?x`%svaM}zmr;ooYL z(wuV95g6`PMkuE2Yw9gti1KM)X{i+uuE9zqJQX1zI@P#8L|BYA%;wJ{_7;^IDAnaLM!VTXhIfavk#{WX%D9heNk#u})Tt#-4NVf)|@!eA^N zfP!uT7y6-^hQ2$nAO|RFeEd0&3&0Qv0G^Wna${RTkOxcwhn1l28oYV_T{2yTnQGxy z5Rx!nHEtUOFTvja{%78GsQ^()ZRW5L_`zNXGz-?_2?QCz*Tfxd0sIZ%i-Jc$V0kL& zclI6R|IFiJwL~Sk9|;REHrC9QQ!T+F4`GKokRRQrW%i_ z$g(u#2Z&(;au7Qd=Mi~=1CI%UhRxMdN0IuXDShG z4TET)Zi>7N^!Z}wm;+2GwuifAj&NbQKj!0J>fRx!=Ji`h4m<^`R{*K;FA-0t;q#}! zH!87lwVG>k5(m5Q49u(KaJucI$7=r?ENS_)a5v26i+xgH-e;s&2xeHI#=cnKcTX@L zt{UH=8N|N5?<*nEn@lu#zDuU=6xTIF{#S9f%MZP1bfd1LQ`z72KDz;nO_n7eaFu{wZ7Qshvj}YgcO&j$zVc z&?ANHnDWFuyESg|$zR;VhOLgF$ws8jDI+9-O>Q@%H)5Q@rif6S zP~p}h&+sSo_^*Jhyh(S|4|9~#-FuWm7^^pL-T*_$)5GNq+>uFt$SX1)kiB4MKOe#r zeOO{uE5>!}1UJjadyoQ|jq(BOMt~dzVix)zSd?F@t*9?=LI0)+C`CM6D)$2jIOF5vK&`ntI|HuUNAPXyGoC+H|F3@8Q2`=0(}J3sniXJl zu=$z=UVG-00H{a*8|i{rJArc7veHZ_3=R>(uHWOwPIr`d=v!|pn(fANZ1sW-zUL-6 zvNk(@>55Fie*V2AQT-&mz&PBu4k=Ldt$6;1*Hs}T24^+afwgSjC|;p?#wN42m!ND> z(-308(n(KR?#lot8;57t{BwDMWqI~QhCO_d4!sosMvdecp|rJ`$Rp+sh)})BO9Qij z`-6xfm}kNi3hjTZ03-7H%lPhKcKou> zu^{$$LOpl{P)|ku6U-@guCX1{L*7Im;7#az5s@)gHafQ|TMe(v@wBS*AR4q12J=?V ztoT_R>BqwiG|W~S?heSw-)7$i)Y{gOj~F#;7er^-`lN@b~E?nEImx7i^f+;4w; zz0bC@*pA9OV_kq`sK;p+>SR&ww}q9h#)k|U296vmn~YZiIZ{H&B5u3CzWus{)yXgx z_2F3#)rE(0UZh^K6!((AO2A#q6ezK%@h0e%vpai{du0$i2e0;SaB1d#w;6oBAvzgG35XC`LjD|uuxDJztF4YmX=rxh zy?4)6khXncC@*kfc^%Dx>Lv)SgrsD6^Jny|&9${|FtLP$gn%pR&M?OI&cQSI94?IHobXZt@GyiA8n;?bCHW(Ua)( z)cLFgQ$s3P?W&V!()#yXhLje!@`C9^Vgwhi-74R_;?vO9DXFs*2vwONTqoges@rTH zIXfJw*_^7AypH1=FfCobBp;zY>%Ou$;E*9Q_8@nYnY9i(I=n#>Yl&^{1?(eOT-fyt z%>d`+&mATawA04Ns9!%bj2Iz%z(L4K- zPL|b%PGok^6t({|g9X|{rfA6wP(Q+#*wYFN<88=4{3;Ac3t~N?>5Uk3mNM8qW8ZFa zDq1099>Z1rBUhT8OUH<+Kr}4lykqH0f-W<)mZ1ebgrW&-1ST3pe!>0J$~P?+qt=Labp~AJBmSS~Hd?C$#|9OnyS$lE& z@~XF(yb8;Vc$9`kt8&aco0Ef$NtM!i%95TL`q0a5lIj=>&<;+X0Y0mu;M8-G^64}h zXzjn>`2BBP^o%NNbja|)0=WYe@+=Zk(qG`s@$;vliOKs05LE`gt7uP5G&FcWgwk?? zVi3)fqIEAia`0$pX?uwkl#b?xHC^*PBCU?MYtH{e>Pcf4hFx(c>>wF+MBRAPE`Q)Qd!r0_L1X$?Q1L)fH%Dgg9ChRX@dKb)Zbm(~}}C z;A7ZD@D6&oYY&H4yi(;hw04Qm->b++aqDg&yQ#rnaj7-aqu=MWp{a^b;$NXtV5k#F zQ7cVXcFxo5-KdVj+kv)CA21jR^(82zb9mF7tD~@3a8zL)GK6IG0SWWJ4ySAIPsSIl&OT-u)8y1ARC`8+ag~W5={z15re8vv;&C@!3THn& zRNo;E=~!DRwwuE3w}^AAR>yl#ZB8-t?eIlK)a)bZ5_M|O6q1$&VVjU!$`Vw8EsILb z8}rKBE0*wGUCr!D+JsF*njasm zNx(B=CX>Ju|7gSpKO&y5SO+7ub5C$zYvTP2u}LKkBJBzT-EP>&f1kKsM5q?3QV_6{ z0Q#&D*na?hPZ-pGR1^fosa$NpT!HzC7!Pj|)N%mT4z5vU3C5&PK9we{8AOx8bz6xg zR>D*m1T59aVv=(0cy_O0ye0vLDLetAy{t(Ncv1|)W#)lfRSo`zf`$gM*5zlIk9w$2 z<~Y&! zU&wVce|`rHbj?=IW=%qd1~n2#n%@*7gm79{;Axabp(sm2j8Ljw+Qs6u*Pjd6k%-d7 zg-}ZZu@*H3K%w*Bwsto*IxUESV)60ZD;YysLOEq)JIofN@TJj}@p`RcW7(U7tK$=t^YxT;Xn2pR0$r&0_f;z{*xfS z`&(fZVw`?%0Ik<;W`Z00i_B^ZVA(HxPiv+-oAYsg-QUCNgNNHoAenikB` z7sh?Zg)Ag2yw>U{%Lt{zBOzG>z(h6}nCJ=M3iJTQq4m9Y%2}N^Ru1MW-vkHTTPp{ouZKJilw#F#sQ_wNfw+FR1?4{7yFnK_6Gr_?4%b zE$gH}!$sfgOuTecUDDcu)s2@Zy0&cX9%XlW(Q=)z&OqW#%ec3VBE0@O)iMsRRv53f zi;wY%+uvEtNn#1yzhGIh`7TvREZ&R|T@EDD^AD*2C2J6EY7i2#(|t);r>m8Unt*RpvkV|4sr1}XhEnd5t< zX2@5oEB%eGO{Ha14-?j2T`~dTBtxKomH6dw&cJk07kp3=4x1dd^=oU4Vx&aBShk0S zUH&_234D{~Jt)8QsbyC|Qn=l5vC8(D23RQ#|H7|&&8wzsJ5ZUvv>}gbBaoGwLbRe^ z%n+c6hcV5Auf@M!kX@Hj6cgUCFFAs@LuJ^3DR53d`qrVz9u{@rj*58cgY@s0s?zby zrh6v@Oz%IX5>&Bau*~J5XWy6L7FX@Rt7k-i-MdP}B8#znM>4hf1 zT$*f(;hdXK9$TqJ>*3%^8KehGe=7nx?7nC-dp!#x9lan!;&$x@4V#1&BM4T6C=W)! znVl|FX#&ZYkbnTmOYv44`!zO07N5I?gTpbHw`<_+7W@vOMS!;Xo`{o%lJafFM__gW zJw>Vp;fO=PJON&Hz!g8Swnm~YB_qRR(lfTT)jy^U4~5MKiYg2Sg0PSfbPkOujBbFu z@#&eYv7}I5^nVQ)GP+{t|ITHW2Gf2rN$b6d^nt0V9B#PR80hE-wq15^Uy$6k_0ewah@<6%JasF^TxXyw$E#F_^ z84v@0b^TWC=$ievabf3?fA>7FaFgPFY{9N!v(H7>ABU)Y)oqETN5FwT?Z77-4tmyO zO+wOovRJ-WC?ZMwhm@;G@;6qaG-yr&Nm7m7UfM7WS?nfT7K63EgGL}>L!>ZXiFN}c z5C{rkkq8-OhRMCB*;mQyZvvM#Sj}A+H6DLY_WV@QT+XbA(gr81S|GHHL{~aG;65er zH==V!9x~-&P_Ct1K34r2uS#0V_NNoELK0pAL*@31Of?b5+(U0H)T%W~yz#nEJA~b5 zn!xG?G=*B0M|QmI$YX{jVm9tLYnnYt%PhZv_bsr@`w}fk1~S$hE;)$WJmlTm?-{%> z7(dq`duo=+Vaix8RpkG2H?GC$)~LWH9bUj+>MwDy#BxChUeu@}?KTbRWh(_4x-rtz=m#)SM^%q;*Rzc(3@?RX!$PdF&E+M zK@T%hUQ?B%)7gxhu}Hc`R+c^kkGUZ8G&6N_oWr%4s8D!XkSZ{zt@nhx0aJbZX24Id z>^lOTHV;g_&}U-^7lId_8%QtZWM=Bx+tb(GYm}L(fD}$Z01XWdtP&g!5)xAX{b{!r zYBI=(h&KWtEDU7?VI{$EA-oZyR__TgY-hYcFGIejaY$4P>b4qF4-RX81uatnXIN?hOv@}RY7wHm(^I(KNS}uB`doHYB^HOo#iymsmN~`o z3*w`lF2EP)y2NfU-!lBDBz&1w*~3wRm#u%eSi9nAq9vFH zT|c82k50qsdI5S2)ZNJaxBSxnTS|&n&{B5Lmg!3U6fvI|ph>UCs>kr2wo38}- z#1|w))JWollyLF!q&4K{=~$;?Fx+3M3cSNDBCZbg3r415k;DP)ZIr#n47$32m^beX zr-m{9V0-G9XyAIP<+OLdo2%#ukgOG()Z{y4aIMjBR@PTz+^U|H!j6XFDuA0}#lkJ?mn>0P=L;YL2k-}!**d(DPU6$f(68V_e zym6IKUQ^yr#KZT2TSFucM=r0nBxyquoX< zX1+9q`(SwuoIG|Q2mKn-BZCS5QBG0uM;_t7&39ac4H%~e80=8_B>f4il|=u&4pkbX_l}tdKM#91x%8JcS}ycJah2Q++!ru}GownR z*}sUTr3bm^fRxz+7+7@!O;rtD-V?H`56XirT2#Yj-~O}dt!uc@ff@w_FGaZEvW}d) z?geHUK@@tq4yemx*R+TU<5^PHSdn}T|ArK~EQ;eE)cSRc7tUqipB zn>{eRIO;nJ@=Tb7!jI6|QlazQX!aHdSbKSoN(3mTB3O7KLC}LxuD$_w%fBcwFj-at zM<#HI`Fb<<)xQ>fM73Hh(5Yy#ze!%x6--_FRV^AU_J5y>-xC<(#u-D)?6(Si_VELc zYl$iRVTdmuHm*gPda0&SY~W>bUyK>riB^9sI6?h3a%|U=c0y0j9LnSEdw`+q%%Bkx z-_P(&D@>HWS3xuXSwExNkF|Sub^8as;!?-&aGtBSiuNozm|B~zk0N?7b|O&`rzp!x zOw2?m+N;wo)@~uxx7>^y8u;V%b+LEt_O@7#%u8ps+KSECkxjVu@KE?_0Y$X^T7CYr zxZ>q%H@8OXLgH3>{&#xGXu}_~?I6L}kUKQf(qdbt@Lq4 zuX1%N5@`haR%|c#Hx;SP_HT1?$6NMwoT1~Y0$(t0ygwlnrpuK|%A3+{{fW%RuUpsn zo!G|OvuQK(<$GGsi>`qfMh~0_`1u`vge27?@>k_)AiM`&poa^-Rs)-MN(xomaUWQw zog7uS{r)}LHeu{`Fn??UCm~>efyLno=#}*FxB$bf0@>d{7-;LZVdQ^rodzHPKK8z2 zxE)G`hx&fUPWwH3nrfVU-n67pSFWWJwbAIVg|YR+O?tP>s@y#_t+J{x<6TjwN6qQa zrnu(6Q;{SN0$--teA!uie+*1!blImt5yES(@ZVFDfDvfQZT3Y=N=CM$G5_BIzI>0= zyb$Be#T@sb^pU(evuPb^N^?)Q8o5s?@vZNJzgWrUhN3rp+qa32kz05AEVbG>t?+?s ze;(ynB?B$4dE_HyUCBED5Mb5#w9*vmbK(S>56jLgVY5S-4hLSOX5|Hf_)gltQKf@l?@E4U}xVxsQTTbk7cyTe4}M<^54@H zxpeoxxTJyp!a_6fJf0E0Uf&ssXJlYlnOufFBxm?9st5!~3~I$lep9}=r@D}7TW-Iw z&iPpLq!1EsT<9U8(xy}^RDW_4n}WRZP9N+DYvY)*tli6-_&xIUegQjH9fD_x74Ols zxX2ur0pXFthECy#vd!^VCzmvYsVWKsbt8n<!eAPY0>9$H9UWLPe z>@UfZS28M!+tPDN3v^M)-X#gK&XW^64&AV8BPlw;i(m01(IuUk>0Op|37neKW#;Z-bI1a~lHz6i)$9{ltvA?qxy*n&{LZLtc2bDs{;Ia)W zlRJ5Fg-Q57c{6;7v5)+%pO0VLv*m}w6cME$ytFThL`WqifBoYU=6^+M(c2EvXxg{8 z3RG~u39Yi$AyW@dcV@3}`?VBUScQ+u4z1`m(_#LqSym?30}EaqO_v!;Ol)rWwt+B*4BMK^Wq)1QN1bonWwC#Kt;sjBN4=(iDI=kS zplci2dO<%CyfPhKBh%~r)qznBXXQ={zBEWD-p_#_hJOSC34_Q>iEDjZ)oUDj}hR%yhUv?sqI@Un}ju1uYuC>8 zS$177I(4kuXW4-CFup(KSbo(y^V{1DfFMm~uFZQhVKVS=U7^nrp z>P5q;iCUMiY#*F7|M#dfrJQu1+Z;VOHASk)81H)z$3v|g%K8(m@kgU5>1H4N4QA@AfTzjM4Sh)Vl_ z=IM|YFo8}%l)$D^l_e8YP5Q1kE#F!ijfpg8L0W2%<6TUotJT|3A2n@8!lHB(ScpZBxJ3BdRZLp3h7{sfawWjcnbnLwCg(j}9VrsrOF`+CoqB1Bc zVZU%;BE%Hn6@E$W&ZA?qExnpVPks0(u^w8THz;n%S1I$RxlF0(Ilf(`tOGIOB4{T{ zhFRcx3m-dMGaxai(Avh?(VywDvh8pm3i*uYaL(`y;52KCmP&d~#Q~YK#icUpUAuFE z;0P=FcZ9ttoJ9vX#``}%kpR5?hv(58wU+FMfBwq<{l?HhoFC<;L_R#A2YPvgz+{FR zGUb+Q1*q-Vo9~hxf+T|lCRi+60MYFubAdCT-qeAC7FHtm$EHu}S)6!W57t!1vrm@q zc|NhFuEsJd+h9b!W?#RysYHp6WLhVs&iw`PY9O19!E?t?=}JOesy1G?m4jf%vJ2x; zSY(J-#e3ItsnKET0zMOnm`jf*S`)?OUuYW(XqYlA7y`j_UNQ2RMk0O%EBgcYEk&IG zC2wlKLX=!UiISzMB7X)Cnxp}Yb=z@=z_?U;Q@%qUo%$`k{ZNL~ykUfj%%Y4Vv~&2^ z-cO`F7a9VqeWOUL*8vl|ntUK50z5Dy;?M2s7h{eIZBbG%!lL>O;{)uRATb=D`7*1n zz7qN|zPPOEj^qG$V!OxZ=O*G*cgrEX6j>Bjola|B#YM6OKw%91Gq+-h%|9g;78XXM zlkm8#^Ko&dUAz9fQ^>jg|ILzPN}1-8zQNy#eqx`7lk#5kfK4q4W6h&Wpx_mzw4?hf zES;8-gaG;lldY4s?L$H2tM^gbHIxzU!^)AUKOYIG%=fvH!0Q0t^N6M>~= zIvjs=<*25oO6yN$=_0m9q7i*nRF3NGm91hbW!-I!tcIV>pnbQAW)w$pmEwF>fMfk+ zHp^X(_$aATuYiW7_F7$>*866>w1$M!+RVvnyu-vL(9K6Eh_@#-7DIJgh)K0)qPI=( z04TSAK(oascUzD84x6z(Qe_!9eBSgKt0r{>0vC(dzK7gB+L_@bu#)qrtl3d+We(y> zNnNEEy_lkariLm0pL0wZqot|I<6?LGU-&-^1`gOMZ=`nry}(V+&E9! zaK+Qc1flGGWr%@@ha2>rGyOYY})AO>PuqVYGLUXrC1?cT*WS21|h7aKG}6O?ViL|XhW(P|9n)cSMeooCZd zlcKQKt!UolO9z#c&`U`jqz#;a&?q|=~bz|{|RIJ&QL=5m=umunEyETP#veUF>5r4QPL#6b|Ct$!}k97x7R0YzQAkx8%r$WgupZ? zIVlNrZ;S$^dk*-D*z)klPIch^RS*q$w(y-ySMsdqit#dm2=F^Twy;eQuLM?Bu+-TM z{4G3ih)EU(VF&K+KhuKClPo3^bU}jD{gl+dEf_?sm8NUtCVFF}96qm6PE-J3B(S!W zSL#u4T7S*7C!$Ywpp|mRLf6mV9pXa3dn*>iK$BAUGP?T9n{?Jmh4VA|H_6`+_H3JK zca&_7Hpgz-b)2?^p}VFiTXf4~kilr!7vu|Q2rsCk<6{)F&zBsw2B4OdX{pnmj zHZmb4678IxV_q6VIya1SPd8OYDE$~a?MV8(D$YdqTzfw&GOuAcDZ83-e7(BH>Id()%ht_}9Vv=V9F-1tHOu{3`+fggW3HLm0_8UBN{tre}MD4?!Drpd4e(D|^ zoiQ4Ota4<-X10V!IxE|+^^t-$8S_y<%!6Okl;Am(ZI=$MuKe@ehJ(`cLkf*}xpP1j z%%3z=jkijyM`bRW+T`Em(gc{fOVajVipf>h*gW~rr#F;p1ZgcE`5b6*SE~}~+i0s* zk5H5zdUTQ4p{8ib(b}08MrzWt=y?Jev0cZ7pPALB;*nG6ZlRtEwUANMB#7X9pzO*l z;7*9#YNWc)_YTBuJ++9~>(EpvVMw{C5?k}q^~?KWFGA)0c&HI>mRPas^*#-bj?%KS zHZ(O|C^&bRZi7$&sJDH3dSYj1Cl3mR0823lN3oZplY~A? z{X7WACBX}v3Q?B?g{0OMgP z+61|u8)ImR^^|N%EsB@4(w*);@mly9WyJE#l%i~C)h6BEzhTMy0r+@&FAc<)0$fUF zn_*NPria=z=V_hYov&(BYRuVfp{js+-t&+K@Bnx{42ToPmL`*bWL znQT>vQAl?}*)?S6ylf`e#H1k4ty3eN8+*HEPaXP_M9~LOxpk6XH!ARFkwn8pqoAtZ zCZU%b^-^117!`m6qhLy;z}G}fkIeVI_<__w%gV}o%m7CuQYa2-)XjBlUXR|d~3L?7w-`20RU+`_=XbOYpPx9jh| zRX5YY)1-}D9#?}-kG=6!+Hm&`=gT@+m?i)wP|0E|0ud`XM6|K7vBErl9+(rdfAS1i z4sSnR|N8cWWj*56zl1Ct9I%Q>{Ehsz_WS8d90OzrH$Hie#-%grG#I3(<{AGczl&wAEQv3+NhgJ_2SRR2FdIuGJ|S#Awvi(!BADl!oglrWyWLs^%Rfh4XdGOJ9ZG zTfF@{x|8~5u^x4TRKd`KPW}4RSB$5)PEO`{S(MMi727lR(^dIZH+l#N)Ett<55Ce) zoBn>P!}PaJnF8)#VNwWBJE%V3GY3I~eVa(L&W!f4)OzC zM&j^5zO$-tgoyv_3EW^<9LU|eI2Ymxv!z{GNy33PQ}{4)0z5oCG(2DOELi*TZ)l(elwOpmOT!l-d zsVtOi)Oddu%bp+que;1fQO-!C?*V3+Iz*K~{h@Jsm{#)z79LV=y3)L+xS93p?D@jJ zS|Z+Kr8Av1R|;Lg=MyQ7@Ff0Ty?#8cAsV@hn;bz} z4Y4fU`&Pt!i&35b4nTqbw`SzTh68xm+Y$<7tmswzxl?zjm&ES=i({NH?h!Y2M`f9HyOZS75 z5C7O96w#SJFj$?LoEWFuc)Io;@~um3(k2eo3iZ@ z(lPP`4Hd3FR&o8<{i15$p3+i@B-tbJuRu~2C|dR0_kZNhXB>5Bjy|{4d4T0 zz!MxEUXF+IqcPN9mKuKhQ|LKf5PMi-Nu@BRn0@sV#-Pfab>o88kPBLATuLCaS9womQaP&=)*+sJUKU}DW-r;~PAomX#@MoC8xcGrs{f4P0f`ettLa$V2q)D%i~KeYLpNrxM|m+VPMGI0rNYPzD@YrG5_< zSW9spGZ}nu7m!!WVNQ>{h5ow4~+M(|^;LpjA!==x^8ZSH3UX&)al312A zm6kS&VkW!{kH-sSJZTPO-jaXH%KYf8Dvz;Nha6#l6X>?oiVBmq!qjh=s?X*Uui_aP ztH-Nc-l^U-j*(}l^Hnq+b+uKkk28mKNWY?$-%GXe1xMLPILTxko6||dSEw-^^_`?RO5iDF7XXLZP=+9-Cp)ci-_i}Tm{75?9YP4M8D0$)&yp3ft zBA)%Yv`C^+l6CDgNFL2fg~jqek%f`>}&(y`uUAH-h}o`J;`3ftA5Bk$$3DIBTz*ID-NT>}Gg zYrC)66hCZ5&m~H>gx4BD0jcBTHO-#UODpT&D+LT1;^*`ii*J;ZjA?4LemmfhBd9vA
    oCGqsOWv=PDp!fQHTlLsQ*mHJvtIgB{ zr=x^#sBrULuFEj=MO^Wmf)+dsmQ)fNEMiIwwiX&yx|xF%E<6GX1OEXjRsVIzZ!9u? zTk4oNQ{VpEYKPyCYu(eqUV`rr|F}CYcGajX*7t@6F7x|7yKH#QpjY>(2nn#s$5>D5 z>%<*#bBjgDkvS?{!!_Dht8-aUb@{|)gy|E#cQEL{E?Vl^s6iJ$n~*nO?IPSeI!}=4 zR95HC{Iabqs28e1yR2yUS8=EsdmhK^C5TN*4e;72=(&XcoXVIH)^gh^1?<}#2SSt8xa9aBvDQY{_Qz>WY9g9DcfC^^Al(Arqr%%I*UdQY+ zfBZ*ID+8bM_LFB3_#C~=+&|8354a=JB9Hj-zjfmVpDVrsZ#i&Y^gloT22M5x@bw35bc$SE|ofGtUn*zMK9B09qP!cV*9;_A%+(9@6bJ zMacx7Y+)6ywcf4bO=9D{ThP@mf#LN<9$Lr&O^*P_C##bs0>mE8dW$;K$J27t^scI_ z#Pl5Di>r!eoer zvrs`)bXBQF*mFhjFD@+_C&@M!7J5pcVK(LD9IAnZ)(-NY-1Nlc=KDwm;-}=OU$93N zoye~Rj!3_E!g3IOY`iL6|461D6c&D!YS8^YAli*>T)(DzWm369@#8i(O?B2uGX0^F z+$shB_Z0=D@5D=VkHyW&2`pWQ8EDrwO?}@78rd~}_Z)n7bgf!_Iuh~BaGgfMM5aae z;!qPCELxW2)kwvAyIro7(|vqIfF!ycI+o_|$a63I_K+OyytCy#aC1B;VRs%y`=47fO6jd&^DO@g~8mxhun+jzaR7i1Cur1RlexBsrY|GvBb`M&>o zIrMRPlq}E*_#L+**C-3U@F6X-SZ58Nxy`_8L{bBi|qA>bFq9G)_P!e{oLu?yM zI!Tj>ksqcF(w0So)H9V3b~xx2N4MbBeSXO`)J61GB0K%2kaAsmwRCEF3du2(jDMHR zk(a}YO^qw)!#XRfX^Sk|O$Z8c%Ql-it}Vf|q$O8Qah^HLqx&0eoJDx2lFyDyqvBL) zZyGDRPGwSvZ(*2?+u)<8=MIZG-@AR7)mfLWV0*wtL8z4tk5Y$GrCEvlczg65VaJ%g z9~u(=1TE;kTmksC{G;ooMBjy;R)>uP(^tpO2G$i%)Pk(NiD=R@L9@HaSAr!K>-s-F zSGVS(5%IfAeik2=wYeQ~zE=%(h0S9Yy_WFz1Tl8!)fRf~>NLx21Ox;ERud*7uR6M4 zwRysUKEN&AzV0z8`t(Co;O{TKE76A`QKQhuKY-@|EYyDEXIBS?P4EJ5;wYfD?U&?L zkz7CeCq=s4ysKq@^VawOIo;Ykdk0RFAH2&OKHA^nmeXo3#Nsh5D_B;gL^tO|OJ6rp zqQQN8|BfqA{RLXA=0)+Z>~o6hBU_vsY4wly=6SNzuRJiH>=MUuF4bSc>Qpu};jzW8 z&!HB1jqO0hh=(0BYz;qEOIcYJU+LYt**R6+I}wDvXC-pfQKqFxze6O?wW#1)>W_BugoTYZ(Sy5=bhv-H#9wAkC67s+XA+=r`kt-?Pg>d@)7lOU6~ z|MBk{a3*Wl=+uGy8u%Rwy6;HE^5cB;JslDQ)h)pA_H+q|;md%76#6h4`momjxCZ*@ zCQToYb%ANl5ReUGZ(Ghdne}1wu&9HP1gHb~8F}#f*n=eg+#9AnN+zx*u5KE)>)%dz zAM=Eit_CZ+))}-AfXPNGC3^g!WRa0g+FvlLml_C7RZVKOr@gLO1 z3gO>Vu?Q$ONsHtlN4@cbPddI0gYq@0?fXEucm(2x-;YE9H*p3DvZxn$F5M?>V4mxWz8`?LyKaH1(+4%* zRNKjGy8rwp#S8fZqfB?fxn2FOM75h~uF!fk6R~%#)OneqmDo4%yp^_qHqEYZj#4-P<#vg7kN?|@#gud67*q>v0D_5kBHwC-){iCxyr>;4o ziQWRSTo_jE+A!%>ThY9>`Nt6oQih@(^!~<|8S+VPcld>ex=7<({V&$O)4xr%MNxTO zY8T@%lEF%DR{1$e3vS1?!c(*+5#lEwh&(;Y_YDCfYGPE|$cAQjjg9ItM(D0{HEPI+ zQ`qio{u}pCSi#B4EF41~#P@U%zhW%NKC*nNQL-9*Evp(Cis^AN9HP>}|D)+E1L=I+ zzr|s?yQVu1Q_~#X-QCT!>F(~%X{M*9d%A0Cm}aI8|J(QXJl^txT%VX4H9~Zs zp=D11jXHk9&r1tpEAGox=___rGz)4oNrm9*VJbx>$q6^<>5>L4-qX1V++AUGB$-|E zn=O~!j(z1Cgo_eRkD78A$#?Z6)9Y~5`)y^|eV%T6uHdxR-*P>|dRr`K;aT`~Tg;QU zMz^I3NPF#hm~`s!yfD<(zJDFt@_gNhaXkYx-Pt@&`A^2mRqB-hQT6Q|ZLwXfQYm?3 zRKIb#?y`z>TWkT7v44*zU**Do2#+Mw%n+)~Iqh5$THhgNd=ScN?~>`P^W56%y08QE zDmkld4$I>$M0ef}_df4|U+28C^O)ObzZRX_{)PwP_%_zuXpeimFjYT=A)U`GJm_Dx z{BcdRcXF9(Ll{GLAnLv;J1+OLNWZHNuD#Cwafad_)F>!Bc=$!cR`rP`n_!VhgSmtK zMUqJc>t5@APS|?!J`Tosfgd6=rXN{F-P9qw@f!Jd;3xQ%R&AxZ{pMcDvuo$5Sp;o; z9kQ-1k8F6@5VgcTcb@QAyBXSpIoKcAT!EBOQ1;n;IxGd$x|JzJkEn=p2RUjBia}Lw zC7eIRu$C!9TN`mz`n`%{6sZa|Ym(*k-CmS8C>$c8Qs+$?ch;j~6cu=}0F(CY^fMoQ zgSxAbT0Vu`CKod=5wg@?BP5B;bM&KeT_b0})AZwuL6hhSd&8ye04g)!9a78}@CI`0;st>~Za)C> zPoY(?1%M7fwr@F|pGPwzY%+#*7)$>Ljq^A4bd+bFc0+^n-bmBdgY$R|4e^Mg&b4w5 ztA;I)|9-8?U#$awPVc^)0pa}C?WW6>%0Y$R+}p!MqEfZw+_K@sx(QACQIG}051cf4 zY27qPdO0_HRG~6hBte5FrwwyQT7B{^`BzG3({S?+C1tJ<=~cI|_{7RS!c1aDl7fz1 z(oN>Gt{IV`S9Tbw(MM{mN;fb}$#~Xf5vRX=|At)jQDUR{%LR{f*(o--_thygqoT$JzkbU^^X0hbeOV$<^v{qNOc4%+Nk zrXy1nbEA^-T00r8jTlzayP6p=VBC6|sSfyO>^{Z1)qaM7m8{U4{a}uc$gwci&ge(1 zF=vYrbD;;G??bGy=~C@E-Y7Z!U=o&^82AOVH{ATfT`Zg-<^mM98=1r{CIMOEhh+=O zg@$Nv+fb=1)x$uUnrYTLV64^dsG!HabrIl!JXtdsvn|U7XvV;G?+2T_-x9M5>ca_2Z>x(;*Lwg43A(}}*p&~`*z1MM4cLLXYmQ^}H& z7mVQEz+!vn&9{(LA)4x0{hVA*LaHH3#IVp2MBFj9DXZAfl=Og>4QY<*FC3#Y=A2p9 zG9mTO6Va;f{d~~aOtjuPr9W^@m2pvrN*L3f)fB|VmEmBhp`cJgA~@Gs3|HU9QDZXY zNjySAj1%Ek;OvnlpGrJ%0`OW*hZ$az#x~^9@l}NoBx8H!s44}7Mxvw0`~*H|BwjM& zAW!SW1EMNFGp{Ei(O4@JLu{Je85>qZ-DWSo#$DdQ^jZCwM)o z!J`RSP(1@{*1rPHed&lX@H-I?d|C*kdR)Z-LMNN%o9lYHLq?>WBR`np5FkJnskwXKSTI5zrH04 z-~IP$oT72QMkdaemPMoLp~vTSRH&RxrCm4xTSNAS;Po_2>TqHMFI%?!@W#wvoE;nU1?s zE%{2LNp%jDvVQhq_$H_Ju8lhPXztohtiUM7Y}rZY)Df?cV6#C4Nz9|dZ*b;=J{D=N zo3g{8i!pDo78C@EcGF#S1Q}EWzoa^zNYYU!)~jgHB->}SG?Zq=L~pOKfwa*T%C`@# z&{g1==HeIN)m7+n(-^>mR~f2~R%208^J8M>F(l?MRMkx z5-A0k+2XNXTIlUUwN!o($Wy5v)}>Cjo|4|*X6jWhe6RCushdXEZ}sPYhDGyL*~4kL zya~%`W3*ace=q;}VN#`F%5W#($b{@z>g%c=Z(Xk#=<#xA065&x1JVFstp!q4{}{Xo z?^|ijAvQAs{1TcQ-2%O2q(;^h>4jUfy7>9I>%UAp}@gfnV)|;v@DupCv zhzWCy$l_USipIHq8i>`&0^>&lluDV>P3leB<&x$Lv~pP|>%kvjlyr1bKcJN4O4jO* zciW2it z5}0~(^>px4py~oR0%Ap!UuaB9uAxkOU+BJFuKY{$Chfd?dam`LoRnz`~r`7Eb68|l_=oN#Kh?`LzGgA{cg7mn6agA(ci1Tc!z{FPGM zm}K|T2g%aGYUTzd_#v?uopB-XP|OVW3Ep62>aqbHqBY8ZG^`)a7PbgXIPWscUUs$7 zD!KUF;bJOpbD`l&G*25nN#vu*7LXW)_O)@xBJO+fP?PL9cu>>il}tn@v82Y~3`3Qt z(h&rv-sM#{z+o|2_Syg0n|4~U1OJJJq7-USGY{@J^@$D4|F{v=8?N}!5g2hP7|4~x zczaHiNfl6{yCE@x8Q(t&-b1;I`$@;+z%O8V6|9T8uJ7ua2257+diC7pm)7R9ya=!xo5ez1`a=;l#0dF4PJ#F}S?Z!@P`1B#CNQgNylwkP2-giMR;`lM zflbom*w)?eI%Lm=VZ&n$^P2d-U1mPGLZ>HS zBqV;Qs30lKWU_w9Yq1z^3DL@^-{ZVZ5*_)iH6Q zDWD42ZAL*Hg(R})sf1a_1YXA!K_TXylAfy+DZ&~;R~zILkcTm7TN~O9iHv_K*bPU| z3I?NUfN+aquG3B;+!=)xFp`W&~I(D-2mQRaZwN z8I+QOm{8c}z8Z}lKe~A$0(^a1KJ&a39i&X7YB$pkIT;P}w-V>6p!8~M?X8j?GS%XQ zjR?ON@T#|(&i^8p;W%471o)U~s=nV(C!Ya?MAN{~$@t$g*A82Y!n{{5;_3Eo4fF7u z)Ym`4lme2u;i~a6mii2nq4p7aHhz}HetCn_p7wRoO++yi4MSJqAgDm6ataeLB27}J z3hrzSOp_h)btXE2@yDpBA`|cA8D%_U6nXqu@K}xtw`~^!SG~*;PQTUdGGWnS#$`Bv zPL!c^fOFy=th7fQ1x zsx-s04Og~daG4~%&25%A92B&uyCoH-jzNPHqg4CKLk?(p&=3?6MCcNY3w4LyfpC!? zC_Uv$BYz|Bt)to}p>kBq9y(GVTa)@%*WgA!X{@8m8D$O7sJGkXf7TQ%SOf} zSl4%biazMyEpvz%Q8Z9L>EqVQsNQIsOu_zLbIs#&X*;pt7q5JI+fF+Z`q~!=`gv*r*SS!`BK#zdFO}x-Yc z(`%)j%*J|?YQ?v`Dho66<4}v$Y2;B3xlPja9Nb=VR7J0=uy^@B`QF3&J8dCLlU__DxP!1=qLz>P@xX~ER8?@-jajeZq3y7cPY#`Jqo zq;*E4Rtyi^5J6fjp3|mi-KB*BHE(S1OB->x!t|2PF}!EIW+#mVucMaKi1$*VAzmJn z7LOB|;*%rdBY3wpn`DQ@J=q&jAA!3MsR6GFJ{96w23wO-B}^#$3?VH5&-evdthB8} zy(p!Cf$>r}cf5Oey|uj=*R1$s-y_VrRS)}mF$pe|D$f4)x6N$x_8J0Hk-Ahq`e`A7}DK=Hqg*tLYIRxF639(ddL_yx7fV1Tvg1-ho5Y zh-}&C*jyD3Ao;E>fv?)34Wbl(z~eC(ngCkRMhi(|aWF$VQtDt1WMPoeWq+Zh-IMFk zwaexr=imrqE;ia5Q+xgg8%iy&uT!<}`|lL)>-aDdG$O~tCu6gUQ`);sl_flBvg$DT z#Jf8R&@i*opATr^-1l){aDMf;On@&kQni@&Xm0Er&L{C=Mr;`+#=P#F0Q3Mw~i z&+CUlY;SfDiZ(L`&h^Y#&8Nb4lj!e<+CAhAk{aIDuW>PGiLBnU-rl;%E+eRa1g{%G zrQlqGQv_IJ6r@<(WZX!d70&SqL+HZBdMH}s^OEI%8s17mJlokW8nyy6w)ZPX#NWtF zpDEdt{RYOSb-LT73V3u|F4%oHQ1H^HT~>Vxzb2~Uj2rd8k73qrt^nA7%LbkXfN2Ea zezv38Q>r>#+=m4MM{6pcm`${&#vFRr)7I&`t@KFmZXr zt9;S(nPoiY`ZCedXU1T@AveZ2k;YIJYX2$kPsKNo;^;M(6bSVIQ7N`a``p4R4X_F+54G7e>BXaZ;~1i-3rw8z%Nr!o<=HgKA{V?thP~ zZG|IAtWQtxxG!!)O@4fUH-#OzDix2V$u10kKGx0zK zI-6GAMxN*9N5I@Ig-$7b`wP%lQP|%4>JWytz`JY)xoAWt?8CR}Ui&ljB50U}gmK!2 zOV(f7Jz2^fr(HQb60$wQH=0d3fSAluj9F#|i7zuQ7DTz4K$LI9DJjM-=?;E^S47<7 zG?y$Xoytay_yEP(s!|SZxhC@{0i%hIm|~!Whh@7uhrb^8_61t7R{riC?j?tw(ZzM}7XkN4){``N{uO_mj()=8jJ%MeHf~ z0xMN&i)(T3&`7a?G%9(HI%@C$V&XSdmkh>>ImFdu>In6Wig235z=>UW+nHY4dqth8 zJg^B-vHnfYo`YBZ(Sm_T!|1pUaS;|17&Px;ie%IWn|7`DRR^j|c>}~>)S6v?$$XK7 zF% zMA@CTf@O*x6Hn@p4Qo=(n%c1y9%+=XuiWf_t{;Tu#jF?|j`5klD*H?#f{AkDJh@{~ zLh370h?tab_6>4;PEx~Z+UGmP})q{y6jLr?G9#glO@*DKd_2fV5?E-L`wRPLpE z8(}8%_`d=6Gt2s}lK}SI5%Btw{P+)*@d)$?(6jQm9dv=}BK;rYf(XO%q@A6snl7?z zV~W`PCFE1w^fHRui)4KTwQQb3g-uS ztdty%9W-h9{nHsLNOrPA+o6WBh~%GjvV1Z^Q&$+s`cZ{8R0A%dgZB97_7MIDOb?19 z;$5~gF3Hab?FhZmU(>sY_CiNi;(A)NMsPn^9?247S7?i~q1cN~c#Y)k5pgyN`0YWh zre&ru+K5CAO4S_RQ2wsn8G#Wo{el_=%jFeK9xB#bChLLQLkMGUbQGd_SPIF)GIHC& zz$YC@WN@Po-%U}#VN@7VlAFA#SiCl7Agsp>H(?LWhp-W#Jan$0l?^WitlPSJ^MUK_G-T5LKn^x{WeQV zGa&JZ6bSi!1`tNz8UqAU9RtYnRKB2I<1#7eaiiUJYOC|2d)?2=8I(-|)S7~>w`CLu zQg}mVUCKF~?3uTs?7UqvhD-zxBY2p{UatowmSDELN%KZmp`iuV(Lp<@N^k35- zUbY(EgdemhNz9t`sbd+9^RdvdH7QxKe9H0Kr7|XDv&q=$%0r*w@{NB{gk_>G1u}E?(*tr2~8~Wr*$?lBOIsz3wv|_m#6zFxvN`9bkHhq zB8x48SR9J+0G_uvuTEfZWGoqcZ{XhgKckNgQaF@HCY@M3wp2v7zfd8?+h%d{=n=tQ z9=@{-imv?8Kr+0sv9}&W|JF{i>sPsDr;`u&-}M+5+uK#ote&1fxu2gqJHbmNKYx*O zMj;k#nfo4hUbSw2ze0F|jJB;nD+lyepwyJARs!K&FSp-%9F_rJ=E)p3!!GScZ*!o$ z0Jd{LiKqDA9FM?B$2g|i{JptooBT{ft9ssTd%YRfPanIe^b4j!I~G5!~fK3L+Y4i65v9KQ!IcS8$osO@Z5? z?Gz9Cs&rrw_o0GDz>h$irYA|2jgh{Rj8sGVIUX9{DV5a5F-kIrgWRgIWl7u zSW0AnHhmNf4wB;NPhdCy2M3;4J~35X47XaIaf1yJ6|))_QiYx0iI@L9FVD(nuZs;o zNH}kUG=-`mw^TxgN~+aXQl5^Qk>GtkW!9ALXn!~OS8EFh<;|m|8lGHfey$gC6xjY0 zp?I32^TUqIUafoF?|F-)ozY6`zIS`5)gnOL3of--c?0Xs7azD_@oYE`03lSL0D1sc z&mr)6@BH_9mtKgtj$`f~9zO=dn$fac3RBk1o0Gp{?|ho_rsXj+?Or`h9+y$UBI61E zIOaj?7aUMe(8kr+m^E|;mK{q#FEUyS%`!otF%6rk`KN06Nsr153dj-%Wy82;B3Jbh zjZ#8^zhOv=ez=_zPsmJZ?@gFytlBa@II`6eN{nR0jg+CNViZk~zlO$dan1rE2B5YAq769=+HtFIxNDv%6s6_&A|79;}dJKSueu;Wu@8q^@wY~O5kx1 zxhdU#ltth;l%ctm{Vsx?U6ijFt$+tuJs10lZAA zfop%f)ExSe{v7ivH3tAE{;$f&6@HsX<2;nUx5%%~7kSd-<$HVDXZYXyrGMkWIHZ1G ztTh?DqFu}xrEi^5T*Qz$MU;3cMgt+RNH)2N;pR`Ru_Q+6=j4%Dl_^SmA)iLr*HCMV zA2!!=PZrRc+-J1^Oo?q8mPkwh1E1;6O3p4EGgw+MhG+7@RqbCT1B)AlCW3MIms2cf z%PeNgU+_xp4pQHTdZjZD30wkq9vY5x-2id`3{_5 zAJU^lRIO0B8wWuNxAl^KVn#SC&qM#HOmBsPp<7iVl~NklUwz@9Jx#T1U4MZY(Xmo} zL^#Q;m!FB}qEx|pPDZZFRBcCX8olFJK78Ki>w+GR>Dlv#>Z^mg(bcw_-c{G%O`r~Q z{P6_9SMZIzx~OFYS76roul zY=9PA!2B#YV@5maG}Bnp^X_v6-3-H8Is2W>pc=0M1+>~`Fo9ZJ;(Dm7Gi$2-NIwXQ zkmtRAA0xF)Pw_MkF=V>XdOCkU#fJ;aR&b0S!Q8vB`Vxk*F4imRaWjB;Gh=rTc~53>B3cIZ zstlNPULGY{j$IcjQwQm*S8~spaV2*3ra14o&G8;8*`M)M$y18&T2hu`L7yab(S>I7 zmP)F+W=cnFit~0(BqIID8mk}T_eh~Kjy*UN5OV$VDAgelaHEh`wY9aIhhA>K@;TAC zGibiPz=osaA=qt-XoucURu%y%Nns|cgTiMvFe8WzjA|8y^Ssya z3vyQ}9IcEYt%$2RO*Jzv)J#6m}JWcQJBz^qiB>XcG;i^S~lu2|Ju^D7AcgIpnQ_| z8mD39gMyTQrodhdC6vuQ+sVD2R$btaCnpqD?L?8^Xr{gP6R~taL~joW_*cW<u#Y zx~NI|c}lxksQj|xDYlu<>aHT;H_Cf{vz_AwE4|ygsnr&ME(oAMM#S9bOI7Ya`FBPU zKPcrt3dj}8IWquM&X;H4=FzqjPJFyn%MbW)GlhzlVIKm?Koi+qCrF=!|H{3sxeXxQ z^Pn=6NrB;g0?HXwU5}cBg8o0V#~VLcN5s*0XUvDc<__DhNh*-o3e$+~2`94PkqclH zmX2x$L%Dd1(Os+rxk?>eCtlO3^6Es{H>hzix~BLCA%s(474|-o@&6!D-J*)2?K(&l zKBoVgY#`$tR*1yG2<1?gwB0B`EnATS1%bVsXBK)7Me7!=_*&%{eIUoej(G^vVS$%R ztc-oqpShC}t|G~hDIj4&FczvvMGmRi=!$0XX$_5vRi?`hD;dQJJu@!}9a^Z%(jaI3 zwUjif06pwFAHlqBzCK97Sv)VI8JZhxq}lUqT=p0e?)-Dem|DnfblflP$s{=m#I0%9 zOV3zhq(o6W=LDVTf-iq3)M~0`@>OfI`wu#oQ^&8(&5%vK7VnIce{M_B?{wWNoIRh< z*}ii-S)3o7N#(9d6zqGRZeE11SAFy9dRlm>R?L6gU3co4#5zpYjqXRCVbZLz1~?1= zp$IVcYX`{rOgq4B-2~c@L5oJDWN$$RiR}o zuLqLTsyLGylUJ5s>ISkqAZd``^k~P@>?CpR@moA-3&|H98LCVsoTvw=U$3p+b7+YT#8x_^T*QD=fXp%zH5UG|8mJjk<%L8d{KSV{v z#D`Ym1Ar`bGQ45 zmF$*!rQTk^f9n+}aw0hpA)W`0RP7?8EizPx?YMmx}QY}y4g-qS2 zrr4%(0a6#1&f75-C_G)SFHg9P>hwn4|K7d#bp~iA$y@YgMX@G+ltbvQVjQoOT7`!~r<6(uG2B7!!|s z+@>==ZumXy{bF7G8GhK&+FrKv=>A>19}XjOFYX_c_N>H{5pEINJ|fwf#xtBd12Wz8~JZ za6|1TfMRt{kKI!**AA@&HMK3euHx3w;&Q>&iY2Y_?81L0n3^&iH=568_Pndt1!qsz za;i6Uzc^lqUVM~&?ytVs&~IqRXeRu03Q_y->ixJ~mq}N`iN_T1( zF^hbdL<`MC@prVd^uOX11%xpua7&$RU_oUFJEjh%7YXRJG6wgH&JGy<Zj4c|8h> zwi_G1I*UAt00P4uRz*%&#qCCEhkZc_RVG|)eOQXoz_wjUfwPgWbw$coi|9z*!`XXW zk1MZULJhS^ay?3cDK1%kK6-~3tH7z8K!od!&2+BP85w?WN16&TuwEx@W~f?WJ!z1} zkU?}hY+v|WT5B1jwy=hbWLt}DyKD+pJ0x{_2tV_KM`ts7=#J(hz725Q8=vGSt$y(u zZUm6E3`QTTxOciYWd97Cr1Cp%)2mp>E`;pHh5>qH{>L0YBq-~rw)i%?s-0Q zD)@X{3HW|@x@F-o7SZI1Y1#>|mT9iXrGK~4#mci#&a&u<2Y9u>0LH!U*jelQaCOo1`U0rL>PXI|9Xqr} zGKHV7=YeBGvkM%ib%(+(K$LjPJO+ZyI^B+Tg{FjnZ!1|y9<8-hhb$b8Ig)nXoUfD^ zjhufvoBay8>ptrc*aC*Xh&4Bx`G-`AuA4ufiy6_hFfpS&hwb@z+qm!nMwe&M@4*RTE@cI`5Yxklc^+SuOWQiq9f zV*Kq&!$k7^lIJ|xm10)Zv{59GVZIz%$QThwBoeuJ>&qt%P5fDA8f}l?=y*}bI{i`v zZmh7_9oUAjznerqpeCuTAem}b9 zx_jEr9K_l%WN}s^l>qR$?y~LbDvf$xl4;~f3c6h)&G>@q0PkU-b1Ak5BQ*2xyZ{&^ z@UmW%0P7K8iV5hCeE`>s`79MczU&T6;gAS)r}cKvpXq{jSG(CK$LEVfi(IQ%LNpyurra%lQZW@ zC}Kd6NgYV|Of_<*99MG;MN8@Yiy-yhq-g{9Nr~dRmXc$4Osazm1$LehdWr`Q!-xL7yK?pxs@~+8d zn9}BkHS_1+k{r@wVl6g}3CaG;mVzwmrwwE6+W)?Zp}vW%&j9_HjU5fIDJ`Ja8>OW| zW^rBIA8OINRmkjap5EShub3N=qoyODN~g+I27!812w5d=`O9D$Njwen`;sV=6Vk+( zWN?-VfEkfZE;Ges(C~>%{1aFv3GYlYip&X$#;>(61}P^kmI*S^X!@|BT-l%xqN&$F zb8_RUz|oth$NvNNY>6X>G_P2;T970$SioFM;QPp>Xl*!EPmw4CADYvY}%6E5sc-3YTLoE1jP{ZQFw_c+>p^SN>?JSXM2@tsl zw7$N-s9e7K5_D0wqJ8+e)#7!$Gc>*UQtb;6BH(v5Mh8SR)&czP`#-;wfM;F;Q%bJT zfCS*k%#Y*B76K>lbJiggnwfuhE}vS_5XUSz#!9%ET!+Py=gavqNk56Yg+Ef6RMgo2uRg4mdmrICStw-dFNnw{Q?2YD*3>D}ytt}GUhtEQ z@$T1;dlVCkUSdM*q(q~9T(R85g2|Zs&?qukVH^P7;zrwVqMhde4>g)N4{^Oru~+>(6Vb=|_&D6$jf1 zsM(>MBv=VQ33zByU~b?0A~^IV9APzIc+dXv#94q9tPf*YpsFWKPW1AbUCrTK;QTHt z%r$Yu@)hSHvVE?aVo(`*5|ks^?!d()MhcT}ttM>D?cg(t_tU0_)@l0WBCZDjU@Vvx zyc(ASfnM9rpU}={HroN*Kgon0{s2^Bu|fa?qp8OT5^~-a0hUc5=M}IweY4yRxLMSa zX~#0a0(Ag4Ct%H2dBY*&MSi@mG`>Sd+@!ur^pe;7fT)T^Hugi@oK4eoZ_;hdUB3+-q1L@mN%VBUdEc zO;)5~XT(juJbm;Aza?6#)X5cOn4(woB#himakZJNm8Mfj zxAERxz2x4(_=V(cI$NLMnsZIe{w{8Z@Q`;21WaetewH_^%u4_y3rjG5~Q;4%6cK z(_2&((6QRD(CAe~h zZ-Q~D!nf`YfMEdQ75;{JX~VGS$ohwhdq=9N1cQ`eQivKbQjWnt55OsKW1xlMhjlFE z@t6jRrK~+GiktBAWp9``Jzni_g$mstKGbQ+mVcW*jLn-B z>Q0nSsQib?2A;7tSis^UKm;Syk{v7-{QWxt)S|V@J+|2WJl<*{u3lrID3spks-=2| zwgITpz8K;vToD2WS$jCe23IKD6pXiZbvD-0>sxH(8nt-rDK*ln$ons2{b z!xy{IrldzW>h_u+zn@CxnNghyZt(L!Iy52h;CoXh@1Q{O6DH~6$ObkNrX{L+RK;6T zN83U#bxhTk1NC--#&f;Ihm5X69Eu#p&cncuG{VsuEga)_uE_VsD)`9Bm6z62QLdUx z4p)a4;~Sp>=0}E=j2dm8OoSQh4X6MZ#g+l7CzqJZR{h~>ACTprzt!!fOvJIr5$6Ao zuBVAm?VK&M!meN53$%cGY;a+Se7Z`>}06Owwgh%RNHw7ps;ND^2zHAL79ub~Co zmwAHtHu{5n3ZI$NMR%*0Wa0*AB|$~dgz`ww;n?B3 zviQ$k7!7q)4814BqK?ZEiT*6jG8EA-EnVYoOCB!5G*9yxI)Fe{jD`Sa;zx=i8SZnH z)Ou8k=m0XRa&h2fIl_sTbC7emL?3rkSEJ&@qGFf=+PMCgT4vZ0=xO@WYDhi1*J|2b zo}r%CDA8D?Wgo`p8Cn6!kUam~l@`5vkHtgEbEjRZ5R9qtND|WF@ruOd<%Bt7 zLguDR5XY~sACXUHZm+M49Naq(Wp`f1av@Vms{9Bzi%wSaz2%t1vY`f zJoNs1VDbN+ZC;q~e(%d!_iRc#OkFY;c!Ow~$g4TlQnzKZ_cq2V6;(^lL#{9g=Rh_T zK_`(E97#wYc~3=71ZN(|XX>ZQbn+MBHRO>6$Z|#e(PQBE>{t*?w&EyBb&h&xK^vQ>NsG2Jc0sFa(scuz%JMLc}0QXx+XvA`cJ#?sH0 zh*-P5x~Df)PWl+;kSCkJd5o`-8;3n;jNBT_$VIfnDNTJ9o#=*yGOq;D^T^Zwrz4sm=i zt^ztZ}A@rk5Ojm^8vmM&MDpO(W$vXUUJMY z^`;=ze&TkU9o!izq0}(2fCHq7B2qJiJeIQ;kHS0)q)4G;&#IykrNcGD(BwRH9?mG; zHekC5Hyr3-ERh)t+nH_{m+;_X0A*7At^kA5p5opfUklH`?U4cmnRWTVD9W^hHGlB8 z+`k*iXWF~-8W}n!G!+UgujU+x8g8#0jp&}KxH(&)FDpZ{P|$UVKmq#;2{0JtG|Y6C zksqw^jTnfRK17M#=r2k6T~|iu!fE?FSl*;%zxhc#6u3x$@MOmU42e|8*9-6tVpOjH z0|Xa<*GS3W376^y^F}lR3UV&Ms>_lpyrY_wj#uZ%7?$QWIKsKkX&yWi=a7`3vyq#} zki@njK!rtTP%}^Mv@a@m9t7#iHyI?~<%#xC;b1qMcm1~%_1!u;nKf|`Ewhf6NNo`ozy^w9*7XSNvaTf*n7xUC zRy?cL))7Bh0|5dG8M*6PT8)z2-^pgy!#R0SNW?xqxJWSJRzH zMwi%XJ@6%$!TQF;ohuy!q(wcC$JO$q3&Jh{KKCDCC>QEPHYb8lJ{e~aOSU=sqkeGw{)3#1GAVU%~j3A9KJy& zS-4mNdVEZ5vQGm70n8khCpR_UQQ{PK7-lG{v}7<@t`;6e9|TN~(tVM<}c zlPd(2TCZ=xpIOjj&A^kIktD*ecq+uxi5?SAxPIrk?D2d~L=N?uB3a0Vo+U`9=#rZ* z--rVFcg9FZhWkNB_CuWqTz`{CT$JyOmAxNKs%MzZS^T-6^X2u{4}!eN6$^R8@4{aw<2z*B{v7cXYj5Al#C7{{A#BznPbKyEmQ6)-X751?~CINB9`!k-1L-tR%(F*J6okgcCByS~`xKOD;NDy#=%5my$2M+b`1OEL1 zJPbcQ%(>Cx%N33!yf*^)SwL(buqM87KfCooYgquxe+39#lZqp1@qKW)ix+9te+494 z=bwSX(%^IFXp8~l7w)tJ&f@j_-2iBX9MKSAS7RDe9hAqDHk@1-Ek2Lg=6<98<0?Fz zT*EvzQO4soER&>AVTs`S^7r&3u-RTJg}z9I0=%Z*>b8u8#X2?Ol?mer5eExsGF#I3 z{S{uOoaS!X<3|NN31`durF6|Uk(|A7i%KQxV}2$2)^V}A`>G-G*2H%jGG3@k+{~8M z)AHfNGD(D;N>J;htw**EWn8(g-ox$(`J3A?;`ej z;3qoJ7L}HNzkBS_WDw}!>V!3DMEWwiv9SS|n*2Mi{_64dC zj#muT4En;?9&%uRDKkgBwxF0%{0)H7oxolL#EP_e21Ss4ww}uU=r+e+0!%M{Pq%)j z4I@A(-aT-ao=u4PRqpJ3YY*VgxcA3?xjrGB;!H06NJsHU`FysulrNGZ z;w+6ig42m-Ztg?YK6I2;wqwnTVnN^$yT?m{dU4McH|UDo)2GtVc>mUBqunw376-@8l9}r&hS?$p0-7KE`2W=h&$7{*N)^+ zgncvuC#X2j=SeK zol#r$rXH2h3$v5b%lR)rMfVz2zn5Bk`N{wUWOM@c3NSJm2G|xSfDxnC>Fx_Ij6o_e zu_!l!9$hf4?^w6*KLH*Julq(X`)hVz&*sag0l!BQIS<&;JCIf>K~6x6MT&xvJ89~Ksf)kkUMO4R&z=2FY1nPD$hC+~5HBl~hfWzanNY$4nXlxUVe2FHvttakkx ztx!7HSl`}xvm$yZQBpFv)G^#i`e5}2aE;MRm^`hrAG~=ZI1iu#K!B(GBT&J#P95!l z!%m&*mls%cX6^dk+3+9Y295gxfNJTnF*i)Y|FLXBL z6JY{?X@D=kehePV!9>=t&m<^!Z$v*JvWkXjX#zMc0DvTME1BDa<= zVWY)o_aVWGIj4u?uBl7#W0;iI?^DA@#}j}6m6~|vL28up&FaruFd%HT$?<^n_3aaT z)@-P^`3~Ib$ngM18r}}J1@l>*e=}3En(XJB;q0u~egfZ(2WmQ9CW#9R9rvWOys0$2 zO>#=EA|j;UkHM4HUwo8~^FZer%-QX^n`f<-VI{S|TFkfI{t zM%_NieB2dblj_}H5i>b2Jm^)LXm}BB>JS<5CJ#aPKX_D%&IcOoqTw*Yo zpz@10g(-<%oCqh=bv;!kAcA3RXQP(shE1ZFJ@8qnyG3wmNUk3eTA~MIo zBeb|EUq{l z11vQGoj>3_HJb+f2yg`isG8HcA5(C~?|1tRuk!Z*rGG|-2!O3NE97vqeZN+1vYdoR zA$XH#Idwl+KHZ)H+F8JMF<$TPc=5;Bf=~UQ5SBI1!%fe@Hp2KPR^61>E4>n1fHzmx zFZfNP^R+KMkMG9!p@acBqi*YVB&k4X{-PGx|M|qahJ-D=Un4(#DDXF#0Sq11m7R@$ zEkPU)!=Nv&5C2Mj!#A>Um9yYP~l(XocXxqn(bnw_Owm+2o+&IjX`D-AI z$vbICOR9?0e59skFdEFO-dfb2_4w>s7~_p3MU*Bp-&<1OkL?i8<4_o^OWsUKTGBy0 z;*d@Od0A4bUOFsa7Lg*L#w4N`pJ2q`JHA^OcT|Fh;CQlwtvrZTI^_dmBueWq1SxUE17^w9g80#6wu`-E!AEocL<>U^(A^!*6$t(0Nuk>q&8bo7aQ3`;e~0vuWfk;j%nGv) z=K6&SeBt4gG+RZnu&D7BQrxD=(qd`ssrAt8^;5P}x4$#Ll1{0?RSDNURL7-fZ4^Q_ zCQ(6;-QW85R&vEXi`lS5FPT|bp$Id+hVt@-+lQpkz>}h8dK{yVLq2$!oGjR}=kn)S z_>CqFi5DaM!L$uGMWIwTds1Pk>5;QE=22vSNxsUo4?Qh@y#RgKyl=xKpxE2b{EW6@ z(a94FeEc5x=zeM3(V>C&M-T)A9^0vxe>3jBpmJP+r50#a9-8(>n1Kbm&vrC}5sFuJ zk{*dJ2D$KV(^l)EiD8egXC`}K5+X&awRgnvde4tqPY@}LH3Edw$^+1hWa+D_V)zkfNQUT|Cu{!H#s&a1}{EGj%{c1;}$oo|V9{afp@I76) zjACMefWnpWEQZ0&z%bHys9CN`Dd_7qo!#xs(W3)YPr!tH9c*Y<=6b1$=@(qqd+!wM z>sPvY+Ryb=wx+|YHS@*AR;>m*K7jdd7ksm@O*_rhb+)Jq{4K~;X9*t+s?Br6v;qtM z7X01MPXb~P7!PI|&2k?DZihaD6xSII8auJ4!>Nc`k)F#D%(rKVZYHA(P}Hn4wFKVx zimSy!dQEvHqJ12sfE(Nq9;y^}&la8Q_R7k64I5Kxl}SKId|kW<1?H{_$w@XcXxC9z zhI1PM3xx;5zb*v#<>ec^rgBHo4YwD>*OB)UA%dj>RH9@`m$B0*lTa;82FU6 zJ#?0mzW8rMt8)8sGctF?eQZw&_}$5?@$IrO1`>%{6(&c|A-qZ3S~JDDQ<84+4bjzz znt*f1geGIlUG-5XU6u-UaJzkfYpPzL-Kglh`r~U$&ikf-e9(R-Qnh z$*pyXp>!q-Ik1-OYT1bevTGZz`cJ-vGTv3WUx2z0Aa;0v&lim>B^%3_FkO|)1VZ)b zM4V>8^784gs`LlV{}!UR*o8DxYryN3mDS@66<7_vCG!PLOF(sf_H~meAZH$FBainQ z^c{=g+bLu9xeb^2`}*J8jYYYRSB+p7JcY?1V!e7Yw5a2>zIQ1H=vC4Mr0wr6!A6+u zqgP(Sqc7~*m8-x+tJuVkyLGwQdu5VD&m3NFd}~oKN9j_gI9dKX2{=4TS&ef&e<-e; zLkY<9ea(|-W`UH!JzFP~$28XpH%~r*O76B(zM@bJU@pmQedEYQ5e;_f(#K!pmYYC} zQu>Ia-@dH1i24Eh($WY44JPMK|Ldq`3jGFpy#}>~S}+mQ1adPk$=nBNMO@FAiWj2= z3oG2WG-sUZEK7#R>R+&W=!vQ*@&8H?AU9~!y%I$4ws^BAfn%m8i=mtrloTp^q0Ti- zlTx{xH|2{!?OTLW=3n|dOn8gcaxhuV$``6vPfeJ#5MB_u13`z!PjJAm8$*hhyj1+! z*e#4RSd=Q)O)yIa^WM73GzvKfch9T>YbR@eXw=5Ki_jX6=<=Ye`3w@>LF`@evHkGc zNc+6zgUEbFpxCm%#d^%DT)yAd-@WQJlP%2ChkM6=mHLX4_<@# z091{@E(|EY!EXr#KH6xh&qQ)}-kOqhlzMzLT$O1D-!~!M6_D~gVXd*!lxpv7*Y0`Bzh$GepAA2JP^PTD-W*8K;J9Vo04?}Dt=X8#ia0|0B$cHV z1>bU7gpuk}@Vbb3*Q?A+ z)frPBcB6s;*EYGOI*wUOF&t*83WtE=1-yWUVXsUW-(+!z6E^i+#ICw3Qb<1mfjo8@ zRQ*dNoiOt6&!!I!OP}!_OPIy(iETV6s9CA|p|Gzu@ph+W%0C^yT4da!+922?i4WqJ zM+pdpSwX`I|EzOWgerxyRE2UGxK}ir?KRCL3CG%fXfZAQxA!j)+ququbs1~=Cmg)gnRsVu_kpQlGn&movMn`GbLjTnHC`I-_&gERDNXss8_r72QO=PMb z-P=jEU_`~`5Xs|jFHp|HImvJKYu=fiKXthMr`cor?kH6EasMcYsyH$(SJ3J)8~p|1 zm%rbpA28FFvJBL%`RtVXvNGmIojwfd^ZUIxCi}qorm`=5XVw-`*c1t{Z$cN$u*Y(E zs^cP*SqWy9}yazu!ggeU2!K^3^{k^LzOQmwl5K-1r zFyw}nq+=_c4U*1N&~aKJV?QA@ypIe-OZ=Q9_d?q6X5wK9V5={0#FQ;U&pu>$pCN@avQP6%aBeXpokfw!2Z%caoEclXT=~fH5Y7I8z=TTnhy!LD11K6yXaVq zM&(cYyFJuHV(ohlHWPKY)$ZPXw+o|90_z?-@d6ki>j2rGjXlypto`TN1xP_x0*zO# zZf!=r+tl~J#GaN*hd^RUApV_Ptod?^c5}gbEQJo8gtr3-ZRPv_S^o0ni^a#e3sDq( z2H1!7^S@+#-V6{?ypW97hU_*s&4$YW*9zKdGayT-5BuQG7I>q8+;a5F`*byqm?xXp z?zwRSsP7J{zx0v}3sRV(SNp|@!Z#r&pEBzv&z^=ypHs%&O4RkuBSC`sn=*deWv*lK zv#a-GpAtCq+plkQBS9GxFj^5U1jdCUnRP#Iw}ZT&)h9QOtP#7M7BSXtzZoa*zynmw zH`DaR=I=S^hMoa${=sLy=XVlQ@&v=LfHUGT4rmK~XM~iB%TTHIOaYV^_x<>&>K4(} zIskB%hLJRy9Q!$p$GN)?cD_)h&a4)~FKhdvqIgaeZp;!#QPo{ZHbu`{p^iFG5~z{R zsVqhC%0VYVre;(H`{!rcF|`U_QOjgH&v>mwA-r%_jwQ3q+ZSS7^7>Gdi*rUyvvW$o2DkuR#?t#K23q$#+4=|`2{<55!SCis(uzW!AYF4U#lEGw|w7I`hl#% zf(024bEQP=mb0K(mi5mxbY_-nK$pGh!#Xly-T2AeP4v@k{ZkhhXW2|<@2$Uix@dE$XLdzq90|F&wu;ki3$PbB6UrrW37{g1Ma zaQa7**M1#4Z`n^C)yA9mTnGq%0kDyXFDRM|!0cIU@p1>(_C9I~d%x&>eT&5Z_Nw#d zXHI?BA@fKfSgO7Qv^-G5hyq0jQQtqKW0T9RXXm-4q@Y?*17nnM@_qjV@kAoA*~axh zs(+qGzDSw>W$P^1d8W=G z@>IIufsKrXp8zAk=2!~T``zJ9FhGKu0+GPRU#*q?69T7tpR6d?`he#bPvbJ2RJlSx zuX4_mD|-uXg8SQ-Q+K$DjiC+vw3@IASu9Y_SUt&HE8=o11}?J{nD_GE@JT*XFj3+( z27NX|J%y_^#~i}C<;G0J99`vy~cEqfSe!M4rGG^AIK({6e* zHB0c(00ACVL;Q(0Jj94KxNMq6E&-@2-GMQ6z-BNa*~kNsstl}^;L)QV%fU*`3 ztj3WU#a;)Sbo;o1cK7|&exuEIiY}mv`7;OX(fIA>_67QGl)+F2NPq9u?E+P@SNG=w ztqzN)YvtcbUHX2sF0!YYfZt9}XyNb=^@C zRe9OI($yp;x960fkh`8BIP?@ptK44f6AQk2cQ^}s|VM0iY z-bD6W5I3Lptw8CC-c~qHccBs0NRUmpT~G7B{aK;3it)vQ?sc&Rai~NlWPDl9?st-( zwc00HxaxU#L%5+M3eLB1#Wfkcv(g5yPw)&^)mlT6lCqfMpuy30fKjqkJ*+y%rx8Y` z5sHg`+yvaYiV@O>itS(E8#`~IZOMf@{_wWncnk&W;%#kRV=yqQ1l@*^ijshv(P6x( zVnJYvqy&40xMr78Uw}V|=5>9Kwl4Ank-&ASfA+;1sg(6}CSoTMU z8!jMjc6$4DV#3|#YlV7o0);3vf-C};LB;R(pd1u4hrn$@Mdcg76d8X9jJXw{T3a9y zDm!T)I^Dbx3m18Pd{_3SKc|g+eC?g1_P08a8QOfhKhg9iC?w_!y6m(`O$ay!-i0dE zioD(t22efe_DmCZtqz-y0+ES1jjx~jBWe*DzUkl0^qW767`BCLTNi)ctE2n`r|CI&i(hF0lz@1JvMq6uJ)u|CtOl1i>j^{jPH_Z+wp#@ zhK&o=toR-4ewdE3s4_N3rq7J^958qpr^VLbNg$B&x~%n`)~>of+83SXGfQZpuNFfv zwl@(^>ky&sERz@eb*S)w9O}!NVrP0KNcl;yXEqXx0^wl&DAOEK?KeKNm}VrdC+I6u zhER(J9j(F@K7x4bsZ}H!TDVqTf7E~uyG?~0W6qBX9f(e%X!i3l`;8a6iD=AXL%t-; zfwsZcHL!EA=9aoIE>j7O|9K4NA%6-yDhdra<+W9&HD5y#FT~%GqDee_PZs>z{m&}! z4FZ#hn^@EPqW_7{9tpUC6_ct4<7g)Jl=Il@o88nHppjIiHAZLKE-tniqxKWr^N&^k zyXV}&UiNF+26}q?tO36#1)?8;mq-`ztl6bJ0rnyjKnGID{g+BUm(4{s8>mUXn|}Be z#j%!s?3TVW$U{);JvTeOJuh_6iXo?9!@kaGJ#(*W!{+@3s7*glz;7bRJHtmVy(kP* z+egUR^;)fF+Xm9&HMS6tEp|13))QY}^cJZ|V=gtmUK@3g?W=?X|D)k(5m1QAuq7D!%WRLkoc;DvBL-P-C}6M&afJK-hfC7YPJ=ii@3*4c8I! zXL`#z=ng>thQ-EzHevwA0}M=fy8sP>w{!yj;`Z-Z;HJ-Wm0vP1qmfXgeuprT!}Z$Miuo@P3WXN<@XNBlacPkj_5dZ4D z;y{3Cmq9bd7bcoB6*4i(dPFk71Nhu}#@Qvpbo4kS!F2llO25>x-^l5SPpV^<8I}`K zmFXE)>EE%nMXk}p+;tkBi1{$2O)SHPAz+GF{q%*Q_QF>AO~~28`xenjW1_c!%`4mi zfon%g@b4FoID?9R(uE(J;&yhUFzCvDh&Ptc(I@f7UQK*(&+hFnRPQ}?B-$fMWBK;o z0tPMz_Vj56NtUlWuJd0k`>6atu2vJ!b-l@w;If3Y2uomovbOo~@}))POX8oGcR0#} zEDrpa+emM46M6r^qoJQ>LS#*jg*u&{4w`>GxbKtUQwY8_J5U2I7I}S_NL#GIj|wO+ zI`v>rGF(npTEV2c2Pi|EzT3mkEIxRQFNrxbjCp=yY%&1o4LhzXSYlVym1G~~=TibZT0+n%vLi@`M^)bUyQKRS$XfA91xC3`RPK_G zfMxkh97N8K( zE*q^r&*~mbkhg7c83!FsrRSE9C#vc9hmWV7287x_jMA*Q_IL!Ylly5Xxd>H^7qD}< z#vbNK=qtD3#v19pR|raY)U~|EbCOg8{N)}v_Gm9FQuOG)%Jd!`PN~(XmE{!3$V#=M z+F)DYY2F_|RbAET3X{De%shVdYTA4^?eMuU{A+!hr4Jg`03dDu{DEIBG7mlWd4rA3 zU&*=Oz?0(U=H`?K+3IV_ls60BE_^Wk13O_Ng=iOOeXj;?pJCiCEJFlMU`8UQ%`^@e zd};yo@V9YTwxvvI%oQ5j$@iZtW4S2>lwR|Eke#7ji8iKOCYfVOv0yqgKwBlYGE!@g zQfWlKVfbS4IBw|2yP}8hurOH9M8*z*gd|lkt(~V9RI1D77a$qMqa{|w)hD&{A_Y15 zp=skhyx$Rn#|}sKX_}}r|I018UAt(B;bp}pi5|g9A&7rFU~X$xcgWjM?ZcYnDBdYF z*A~Y0hgYeQ-ejT$-H71*3TZi}D zRr}vFxq`f)2|^&!yw>KFle4!%H~2yuxoj|B!4_@T#}Sf!Lo&pgL)wl3^40Ee`W)mxARI%bbj$=pXA-5qB@9LCI5?B zomjS+IJ74UGu$`n5lIGW`g%C{jqm}Z1~n2m6xXgR!-s38XGQ?k9FA3x$*6qoiViqyn{*~Sh7#fEJhW`m+AhL*p;1>KvoFp zh1dETAroB)3+Y#ko;(|LqotwjHp7uhv5OpQ2>qy_Djl`;k{n^&irS2}^mDQVCw`D4 zZ6w7`tPtgcWK^_IF!I{f3jat%L&&pa}+fn z6())xN)J>v^4LPnkk4vl3J^9QROYtiW7JZ@)8gGNE@WsUL~VcT1N)Gvw)*MTp}q$s z`AVP_bbGyMD(1E;FRXe?0qBpMoFn=L0(*HE~^wjK4p{8xQyl#rS;mhN)p7U&q&Q*&+BP>}xM>dz%+Yc52r<`#VkEpc&aJV$Yfkl@_pnq6Ah znBb$oB^vtOyz zm04p)U!qq8V?^f_YbH$hOteP7foZ>^+&&tXIc7#vQ zo=!mF6anvLc1@b|72qYgP8CQzj|CJ*;}Gz8hkzs}CdDfvQqsGaKFOC<>Uy39#X zzUPana+N$cZ*RFzh1WsTh!|wHVBG<}bbN9W8yoBJ^9jFbZ3MEWH^U_^0EZkT>%FGOyrJ#f71c~3k!eBcuEqvY6ud9StdMyo`#J1sKcz%Po=Ya^cvVv|bRvDAy?RKs zL8q{znX@ExV!2Nh)o$nt?8L(kElXhj9j%tG+@f{>wMx&n-!7ytH7XL$rl*Tbg-bf3 zYnr&$Hc?=tn~B+m;wrVSF*8>X6jpW^5Ta-})A7l`ji*H7ot#lsxsRD>LFpZDKK5!1 zF^-WHK_GTQ3epl<)#S)-`pH)b-k~~lVp_S@<9wyncgD?8+2bEIiWZbE3TUg!3)?1{ zQkdT#l(68L&+4=`X^13IVs~oj3z>Bu;PEi$y;3`>Jj%waDyT?Ciar$&jo)vhdp(^g zEG+DGx_Se^H-?QCGHz~7Fu&)C3TS9(l$Df#W9|R||Fy0LUz_*;Rm_YT%Wq zAYW1fOqY7;n;>U|83n>UV9*QlS!+{Mm(N8}Vj||Z(j%8hKYxD!v-bV`5Yuy{Es}Yr z`pKuy(ltG!s#Y8I!}6zFvt)rWg!$uiYM1*ML zZR{$c-j9z_J$FBiM>6z}UwN)8k`V669mae%jQe~^ja6lD%wJ60W2taYOB2qUjS}%v zk+1$xn#a-Moq>oo3(KzN1M|D2X4{=KL{CrK%-{B0289F$3~fjBKBi+Kr zyT*krizAc~LXwayDtcq$k)$T&gBW`x)aH^Md85jFH2s**X(Od$jp|lc?62wGboN%^ zChPjl9<%lcYV%aS?6hx+H>iOAYU zS+zDd7mMUsysRJ=e{&oa6GI~8wi^Tk2NooxkXx|+(9qPB2!o$`xckcxgpdUu@9e|` z3msi}85EbktFu!^S{loJU=tuh-Xb*v80EAN{Hw{avCkD@X~2l*y^oKjwN4LS#22`^ zFVQ_cJ*}q z9E^%ow0{mbIDJc`xj*tt6@Gu4wiKa@H^ubbPXuoDn z)02)I&Z8I#A{m;rje0=9>#i#BMhkPq5xeeAu&5>4ZSCw@-+gKwN`O1^S|VcqQJpbw z_=iKU({mx^8t54rySy8<81dF?pk~v%`SzuWSw`Q{){4AV-muzDpP8v;v@=OVJBeSN z%>bJD+HOsDtn`9|J0(5}F`#5~(NUzq{&%m_d)2PHRqQOr#QK@23UgKiyf(Fncb?U} zk}Nchx$FqmP04b;;k3KqZOl7euT&f$S`^LKpFvW?MHq72!QQo=f@@h4M}V zn)h2_35-IVrFfHE1=iTkdxCcAUZuTK$}QW{nkpy;gev=xSyt56J)W1zafX1$?jesiY!+-Zi2iDFH>oKX4FT z->ESkfKUjKZ=YWv8`L{*e596^5`*)2qX3SgV^}lNgN&`M?bHuij|2hd)&A&{t-YQ7 z)Ay&_tyM!0x1F&TonAjfzCYsrvSBq50JhP2{N>+Mg6}*LSu^j+Rv#=C>D0Z7vY1S@ zN%>?R4O;ta^#*lCdB!U;iUxxqCEpKY&)pv{(N&+$4Y`O9JF1DBheDZlU3Fdd=eM&{ ziirK~s{Q^dqWz|hearYyX}oBCVBtQE!=}X5fQorc+;gIn}CH3QVS^)dapN?^GY?Z{vXFZ#AtPDJvUBv>R(*AP{` z0!eB|%-~g@1}0D{S{fPBR8fP79nOn$yRBr z$U~`;pBOAw>h~4sbQ+^MBoY6#baruMh@b~vEcwv#9oqc@&fgK?Y&L-Jr>BYHmLI(ar(=`t94WE|DMTsQ{cwuhkx(-)X)3SsVzABSKR_iT@0c z1VQbeu;UV^uOZtgzJ7lz7#em01!*81O~9%L!Dwr2lyL2z0t&Cd(+Y)F*kOV7B~y@~ zkPz`3lqBDk7CQl#byhaEA#pfgguKx1X~9TL%Jr6a&DGTopax4##oa_ufjk5`HoOj^ zSVYEidVt{#$pt7ZppFyt69ok$zzFlVfC;C#K${2JRE;mb=hTWA_IPN8^(JPFypYfM zkk8jEli{8AW1#>VUImfy7%DlDQ6?tdeCTCRg1Qqp_oy9ugvA#w+oU z79MlGIa2JBJY_R)ow3RT@WXnJbxd^0?Gb#cvN?-CYqQ#u&v!9CS}{Ni?fJ%(q)Xvz z(9VklAE;>+>D;Vj^hhk#7q35sTAqUX_AXWipEs z^%pDz`yzg#p+bY)ZK^|HV^%3rn?{js?(F>7-uHlBfudIu7(c56xdN@W*V+1mEQA~% zpU+}2%5`TrJTej(ZJ?T=*Gw zuxOaEUM;nI9xe1w1qDIkXrc)CKY#wr`R`|ri}OY=`qT!1HGovg#>O@~Gc$;0*Mbuw zt%XFIT(FjO0k+)$dkJqtg9fkp3!>~^pxy-vG&67q%c zd)1Z&@3OB{yC&lWt0HB)ff+T@>XeO zY&_EN@`46=JZvaJb}YOyliRc#%3gXN==E&U;}@dK3bp=aXzqG67V#&W5EW&^YL4nn zmJ&wfDmuZpvPKsT#&l`zL1&K1l}{|iVsMxHTK4B{j;FX*j4HX)N zSE;J!4ksQPjaRT4*JYOp3Ry#^kNt@&=TVrqrY-uNw7RT~c?0H)HZWFJn7*a6#z+Ze z8gywC@jG6Y5op!gYpJDEu6W0M$kvUQW_4rIzB3j*n42Bh#PT0gD36yOD_x__tZZpd;0JsUx50{*-tF$DRIU`;cPBda+emV%uRexB3y5*J-qQc;{F!m>)D|x z#*L>`wt0;-`J%K>zFHim*%rA3+G@1otxE#c#>3&L8= z>{Akrgy@VB?#jx0wtLz3z2j74=8T5(^}dsggbb|Le~{R+nX5~yxuSaL4Ib(E(-;Hx z1URvjrCGMCS6_$UoXLt@#3B$C>bnRSueh9kP_B#{VqA zeKB^O{u*D*qS6;-MDbKSRt0~cEk1q{fZiT~O10Bc1C^+#C=U;Au@4v>hAt~XsjCFXv@&r|1Pgbt7Em9C?t$dOxB69o*?o0|m*Uefn)DNzjD zqLtP99$a>9oh#)q?o9kd7-ji_Z$%hadQlea*cfFLVLob@j@PPTAA?tDhRVj3)MJZ_ zzHb|=!z9PsffJ2Fr!kZ+#B=LOD5J?##7>PV$J+b%N3V*nCnQz|Ir*fdAt@F#>bHAm54I!rp5QAGE)x9Dgz7sWK zu@EeJuBJnu2TEj{rn0iKaihf$7&2~vj5QF6xzXv4*zxC^$l3iCW+0?FW>K@xSDHcB z0!7Z!lGZIfGxMY9djdVQ3_d=-7soT?w%ItK2(m>)MD&v}<3)w_bA(Kp>9dIsz^z5i z!(M@MC+Wo!Vuf+5eR_R8M->(c2?+r~1m|0YAob#gh9n4yz6r{+@hS!;rfIhi7lZ^u z9iC6O^lAhuxZGlw3QBhiNd4@4L@>lQRrM~|!p+3(=o0_hLHiE~TurkA<{w_1d0Q72 zAJTv8?8z!!bwoCptI!VK59(D2lTv=<$85c(ruB7q)HlX&>}xK)vzGj zQHs|?4mf6beXD9K2x+8pXC=jWPE2kzT8WxYbjo4A40()_?7F1QO&t5m?Dg+wdSDAG zO^3>&qL4O^DWCq0fN`UUkkDklC~zUa>kGX5u`v!or%q!<=W&pJXoeA|;bLcZ1`3)W zyV%WF;Ya-jZXGfyptuDCNT1Dqn735i>z}+`fCdvlM5ed4dZI%}l;S*+Zi{4Mad2^w z&7+Yjw;3?#mm3}-VW{130{OFBb70_8Qxi9ulw1PY!D{Dk zV90WKfWZs71>*zAZQ#R^X-D1p}lZ;o)9Ej!3h$^Xqy2dZ0G#V=k83UXgBEh|b5 z2f$gJu8x%KEJ}%;n9Y?Yju9H)dLIc?9d3k~D~zjFWPg?ote|&6IDo|(`d8Awe+!$L24Po~@oGD!&Q4Yz ztIUstk_k|hlrxRTTn{6Son)k|f%IR@wsewr>Y?(ECcFB&^xe0D<fOcY6DvgHu@P&k`E36$C0w?VLdchRz~ox0og^D~0sts4S~U4P#!ySM#dE$~1~B z+Jta3x}JUVlhig^m*bPAv>HB_!ZOv%lyB~8uQjfYe3r)$%vR`US_$SEnz6#!n2#1Yd#NIH=F`M9}5 z1`@>{iPVrXWT@py1Y<)Qu7~4E&2$<>&>#zDYU00vdNj;PCWk2j#0g9sI1zn8B9nLZ z^n9^HPfHsV`Yhw>veDbn!1V3E{TI0ON<`e2Q9V8W9v)YoeMh!*MmUgL%zb4w>2vN z3h0R&`EJr>-f?hz`XKjRuiN>!MdqXTser}yYN|@UIr62lEy;|n3`N)OXTvR-;Gm`t z>oKp?D&biZ_r?ud_te!)3bB(3ZkdM_N$yIO-Q;~ktBJ7pZ0*({m$pW_2yT3m53Jh! zF@ZwHu6433<@-NIU#D|gZt56EXMk4zV${VdRY;06687|P$@KaoGm1v3iE?u+QHeCZ zDWgoQ7IU>_Av4nZPV%&eYWU)$q^8lyI;Knmsa8BXW{gr#Wm*L1)=wG+Rc-d`SPB$r z=XmWg?p9(|#8zsv?sa6IbKNLPM{Hj@BgQP>xFVOcHo?5ha7*r)!_cx)cg4=hroOMv zXh|PWLEBBmuqoBz=oCb0T8SZ%N;kZTLtPX*kZx&mw=(uw$n`j!V3oO>E%$xRtX3^F zH6~CgpdwT`ThhlL0bvau}2)_+fwtcFtrs79WFXatesInf3)BVp2#) zBRPVx0zEywI0~#Sr@#`S2S&w{8laWD2?*5*EDJWJq5e`lWCm^U3k&vL|fBX71>^lq`0%Q=(r{Hw~ILyPmOHn&g2fg;#p8|1M?CzMy zO!yB*m0GqNGnh~9wW%k-M0>)qnCi0fZE>4o*qWI45WW||A>TWTV4cEe_qFUzyW z+T*Pl#?Q;d^s3;xxrrt&SoG=}36JK@Rc?R3b}EuE2RIj@Ut)9j9}sH$sU8Tc2}Iooqf&5imlQF$`(fP5c%Q2YL9vHoc@6mvcB=>ag{vp;dxI%ZsDL zurM*fWKsex9AG%4n8DiE&;a2b=q)OO9ZpV8j)Kwe|Aa?`fQQ8s)9Z6VJ-Hmz)YQZd zI9I^H1q@HYboBrj)GaSBpC@``QCR_c)du?_Tx`tF?BGKG*0gN4xnp0Z`mK98^Z?sr`|=xu$xY9U z!%h+@-fr~`PHMfQ)&*|Fs z0E`f*em*RJ5BO6Cd64vFH<6Hl$x3NL3jT^ht8KduCmCH?SqY>-;k-kzgV(OY(DSZV zyH;0?f;g5npC~7rA@wPVdeCXLeJ2Nx2 zQXmyR`~(Olp=%ck;z~JKq<-`a3^ZspGzLD!cat3y;?!~2L2n<+s2wF;qXK7tDnEnQ zO&>2afp!<@85jQ5{c(sH_pgRz8^8fghqX_DSia6N3nJdsajRqQ%5E*o0a1beHJk9j zv~kf7$xJy>Zr{eQAW(&c{rl(R@eYX8Py!&=kG7MV+S;A%?YF0eeZ?>ngH}|thtTAlI zSmZJFyBGpH@!#n`ZX$>eRl3@73Rp^QT+Y@$EiNr}fB*=J&I3@OK0}m1ouNGgW+wjr z{s-KZ2!+&yD!6YYWn~e*VGm5AGd?$)iVSOE!wL!28b%kHc3RBHgvB75lW<$ggCYW0 zq$f5@l89B6@L4zN)nvRpzW_#PQ%=1hoE|gib1f~rgoK3d?!Upx_5M(|4_5pM80uIB zIm=`7n+L^4N9%X{T!=khBm#)=&A)$X3YevJi+>h-{J~?7sVXUfkMcX;lu%Th03CHu z{3((S<18fzrhUxzVJ|GTOR(KP!WctBWI;vnV*+`YH1KIK*`Om_6gl0N&aw7oV zv@NKrqEcq360c$PJW1@8k;Bh}AQYv-!$B}4aB*>A_q)5h8D`c9g#-0i394T}uW}=d z?fDK$^vglYJv}`IgZ1XsMj8>39GEVua4r^3&bPqK1q{M4FffLOhR~-jd8>NEA<9MP zkV#ZRb}~-0px9W{3IoR>`ps(~ZULIj1Vs$aipWSgYis%{xGxBaD=RBS3XP9O+0>Yd zZ-0Zi7>98eFMw02iap?|^@6EAK=Dh<$}VgMc%3{)B)+))rK&pUfeZEP*RO%@^8Qle z4p=X_R!#AhdLfyE(?LKt@cL>*i)j*lt-*Yt&iE6A+9S0dAwbWC*51<0)W~SH*Z+<) zthA)$RsB#Oj;U!^8Dt7va~_Coo=2z<2ge!cTfnQ53482=h3CaG_^H8FE4oHOMRmpI z2DIiEs03l*;e{Fnzks@7R!)v-uiv|~^&YTveVYCFc(oNt>2I@1_lNz=dBCi zppdA0^lz11USAJIm%P~j>A~8h$ExFBt>HUOC_fKV1 zR8%b31s-m0P%^>P$jZu!+wY>+8q|xaTqi3ywE-$fBnb zOnhYKMqvxUXFvG-^uSyIfDOR~;NjK85q0y{R{l}Jz(AruB82wY-JN@^yZ_A;8akjz!F9FOv+fiS%^R% z9IoCW$QomjtaVTwJIZ%RL1pds_BNncT!WaCla&QRg$>wu1C|yUudN!W;B@PZK4fRd zK&fb$=K)_3$TZu|lqmG}_NMjJ)O?+u#_k8t8bL3nGCyKyvF~~BB8x}<7QwrKxB$G` zk+L>v&;paH&`a&ZzMfJhD1&l~Q&5*04Uf1kVYm%s3h%A7^dP9oz*~-W{Rr^|_fq^2 zSF9Z^e(_Qmz0E!dIwl&pumlb zj|a$+ZWbmsw%U%4tUA9W3~7FDZu`YLa@LcR6W6rElABo-6&19pYi5Lf`j^6+kSW1` zvH+sk_$Cq%D#7Idr(ZxD9b&*hrJ|#=0Ji;Ln79bUO+m^41exiNgV|p|G>CT_m$#Vxd+EsH43F?z zGElMaBvEi%BK@&``<64oUNE?t>&-A|vk;utqgo1{FKE$z5H3Xdd$9F?f1ddqSYD?W z2|m8Quxj*5S)2R&K7eR5k;(C6X$d$Oz+nBJsatv}Q=@rX9WPU)N7XX#>}yA`qQz8s ztCK7v$jZ)+frd6X3L6BMVaH~A7|f4M?0jdQEDGo0350+PTGzbFJ zfO=+9Qc`Yi+?($M12o{fwMekAIQ?%u)L>S4SXmuGc?GZ%00t8x0D();v*hLBad&fz ziH{!&zSZ(|iZ_S9W_^y1*fNM@|G6IX8hmiJ{UZSJ2U2(+P*VWev3$m1tKXcCj*d5& z=ty5*|7AYZN>&d1286Xn5_|(SH3#5Tf#Q%AG1HF3f&gAUh-ZwM%Q*Qekpu0+Sbq@KKzG# zB}bW8Dre{IL*Y@;ZR-Cn^8Y{X*)Sd|Di}QjkovT|5x0bhg03Y zU&EVio~LBWoT11(mU&2uB1w^{5;A58iEWljq(LMz4T{Pf87mdq#-xM>LrD^n$g_6! z{l35Fc#rqL_c@;ayYIB^&-J;k^IYdz>pU-5!iTE;WBjS`e5=xhq1Xbh>#87k~YNQ0xK+UYUrRCy!7b!_sbnUF~n*3JYV_5rkev zH1y%OG_Ns#vWMzf;mh78MzfezCar5mQ20}c){%=?-d zt^K|cNspsqZ=(Ukg4d_ zORQbt7)Iy>W!H(zj4)QOY46NNYd|wP;lj;5{{H?)`%9v})rEd1Jj9v|B+aL!aA?Nb zYXZ>yAZ(MNIL(l98JuC6c64;SZnA9~BcD&+6@KcNZ!mwjFrK0Np?4wH)JVH#W{#k02~z%3oV)K(E$yKCr=BPMooumL7cP_^dU_riotW&eL@F&imz;b}D?rNb zO-~PZg}wnIu6_;S^RM$)J2$hgQ|D1+)!w;%d#1_}wh56fXDtvw0JKVf2D^J6qMP8A z`4@+-5+>kOR?)u?$iQl&0u*@Ixby^SJW%|V9zA2AP3U)V>65PUhdUoWOuL|Sa;cm* z$KbV$2?3W2|EQ%Yd`7qKeAtvuAH#pxIG!{d%J&NEI%E*|zDj>Cfy#>5PawmQ*cryv^*b=e42y|j% zY8q|fLR4OuogD-i#69kJ88z7wFxYbqft4L$R&~|5n4n5+v?*hJDA$PIc?H4Y$@b=F z#n}2TYZ$93CQ=MRO^TOCQ zm|W1ET#2cc&aR@$ROL-PY(&yExTy&4;XbeK(Y}f*7}1)o!7D-)13QT$VGblbjlQ(B z^v)~Fv__5SRin|*%*de5$9+TFwfH~++%)D@79I4ha4Ro&RmBizje>1U5@MXUmjYbphcf3ia}a>v^`h4 zj`A{V<}^I79+)^M^YO~umYh$r+LPI3=hk;g95iNXcithZ_`8dT$6R*s9S4`zOFGe{ zo^3m4pIp&9*4u`C|7EyAWZgTJ$*7!jl8obhMqIMJ%)6OwcIK+*@-WR7idp&dSkH^z z(BwRCrX-ULlqX7W1?1gdu9|Mjv&9T}$9#RAwnf-s&L!K}Sf1Ealma{~@+b-nHtpTJx3)ftFhX^$t05sS zu8EF{hK{b8ZhfAtx#ZBplY%Z1lI4uG2j-d?2z2i24<)>I3z#jlE0HSJpmiTi()DYK zriJwCR3D}$+CDfaAo?1YG~t|r%Ax91lo6$6VI}H+NAB2sd~&vHt4J=-T;H6u;TImZ zxxU0)oMt)#cJyZvYWG^Aw<7&uZ!ZrEE6FtB%_;D93BE9o*z^REBN+B+CMvq)br5N!BHc?K%xOZZW_hK|v zE$w!Sv$h~@`expfEvVnZ6432RFpA2IJ*Qa+R}*pX@%nbH2a6S4Ne~nhA z24Qk8CTC|jZCNy_i!m>fYAo79#QeqGgfvz`bymj2 zM6{04)HpLTSo!dC^?z1OP=;;`^ zs2b&mUZ}s(p@4$bu-(XsXV*md+caTSwvyTwpf@No$ZhcUt5f%73jgJ_%+P#dkT@jl&3l#Ez!Ve3KKy5!aQ^ z+@(Z3;D>Vs$156%_e3pe;{BPTas%kB^Gk&R)y6xkX(qZcP=Hgybl}yi+^Q#}=&2c& z)oL3>1kM{RMB0M`p8q5E5ho={bYpof)Vrx8yS8;*ITyIJhfRl}BV5reS7iov-NN%_ z*0PKTq#wQ;$>GVGliFj>sbe#bF!(}m>$z=IIrh0cRu1!oYumqXl8O6W?!wU>Fd283 zhs*epM~o^1L6(8!R~g2nd)J49WTwQB6UG!*;LYu@{loZpe$}I0hpK%k+>n9I&oQuP zp(ZA7Zf<5~45EZ;Mh@fb1&7efOe|y_an4MX&@oX(pDqp_=l=F{PByFuWp;0`O_a}~ zu{o+Y5h!{_3TG7Zk4+ZMwAc?Y^9PslExY+8E{DfX5YD;TwBA2SGcyWX~c zKRYItHK$k&41#TF)!^aYbW>xGjZJH7D}bE-akIOMFVQ+NN>R9ltPL_>FX_Z^2C%$2n83C5$y{O^x)c0y<@ zt$5UA=O|{0iR#70JaSdL^p~3YA4|l;7S^g`K~gVw?#tY;VMAVi{<(FVYx$=Q5aRgWm|u0e>MJK>#3^FAy^RHr)XPF?$+z3 zEx6XjCB~0Q9##+&N~`?ffUUZrw_68#4iR>=*kK6wZ2Zgdw$^Td!PBA_%mc zM7S<2{0O3giXd8pDE{lNfU)}w9v>uV|L|~wOK@1rby~<+lg8AI;A_EO=tL~1XE%Sd z&doP?c_}T8*nPWYq3N*99Y-_<85tRUCd*R@!iRy~dM?z|)Np@1IQWE(qtDXT=qcqS&faz z#4_+Ub^_iBBno7Bg1eO!r%P?_q`jOfbMwO7+!ny8CCye}cdOl9yEVPMycT7JaAG)V zFNH16xL8>!&CR%M^O$V8cduL7c7?gav2UXLHDO|{GuGX=nQ)f57(6z^|1K zH+64HrsVuMCm8lxSe#CDyv<8PHuhSxX3ao;(B?=ZX$!WdrqStXA9xmgd>-OTG6dsy zp07JP;MM)&A@8=*9DrJPQ`Tt1!{O25v^$%2^`kilnej!|<{u+XM9;iIjSEe?HCZ$$ z1zHS4y^tN#$_BG!H@zOo1GYp4?_@gPMf>f1fr=d4*IMpTW;wk_Z^NCz(MP>w?%~e* z0Rk)by!u3a|9M8z_ix{tEN^d&H{D8V5m#0|hN)HIUpiqh-X z_f2^U?(=&zHa=d35oZvhM;38w$OCOIsz4Bq1j^LsAZGGJogXV0FjD-@dJP=Yn|C)9 zMqbo$QHv+fo}J}jadj{IC1D%G+>!j)cjQ*lk}Wf}dj{>rt5j%7=x)KBQ3ROd$B#P%w6N6ES%^mIu*tR_94RO%nL0IRx8A&! ziGx>kGe9Zqj@=d$<{iu2S#vM8#5nA)FPHlw^`Y>x^26&#NJYE+SsNq858bspyXn*N zzO&*jSzWujWdk(#v6h;?`8r>KH>N$w7PxT{?Oz^Yae|2C6EH-^8XE8F4NU(u<(%K)VE;Md9V7d1yX? zpL#vOBSGq^3I4J_ENm%YTQyyytqq0$td=ANEZ(rWGjnajiv#gdSy@?CRT|3O>@sck zKH~Jg#>NJO-#SaTRWTCX zDKsDV4Ge^!d?FloavBDAvok}LQGI1;9(O18K24KmW@1A7my06s3Xl>ztO}WQtOAgG zd0QE2s!-DP-+rCFx;%~nh#Ydd$EO{3?%e6@e(acXwNMR4RK+-;gcm|T8UFEZYwN}v z41|J*ph5T=U&GX6^#&{u{9Oi5MLl?QT$*Vvn1m+aPLy3%v^E9T2pPqGWq-fFu;)Hx z`5N{ERorFe1XkULxK@F}`}_dDW9F43)M#sJ9y)SFVx7(4%{?1MDSn_YQ&V`jix)2* z_Vs0d9QO+02(m)#&YYQ-PLXHVq)eX`)BraHf2<;``RLq;>ZE#uPNB-m${Elj=@8DE zRGbIFRWn3DTv}cZAD9}fc;4352EBt8KS`9eYm zn3o|$1PEU*K2O~6qaio;o1EsYz`_IkH1>ea#N^+XZ{BPfdCaOf4c3o9)_)64B|a{$ z%V`NcGVi{AW$y<4Q^q>8PIHZF zq0NwRVEo)&qk7U9rC@vihsUBEj#^Ix!hwC+J4bV>Vo5QSv7OK;@7UdS#XK)Cd&r)T zqu`gUtj-(a9c26Q_A4>nR1dF95W--F6!n7jd9LXdLrZ(RoQzE3&2!E`5trrd z`{R+rhjUHiU;o)TSo>tGJ$I&UBvV4}Q0Dd>qlzeT)zs?!MydHy zg+pZ~z84Z!K(oVk7J<0b#?ZZ$=v3zA z?|=97>F0<8drKX&($b9XVEJ_6;|ytGx~usq2qb~u0=OrfvAX*D(H(@(8*DdVmYzt93J!ew^vN@F zP}l@@Jbudq6zX`l^q(bYQ|i55-NUC(Lpa9!nE%1UW{fc)+ZweiwtYDX;QQx$*X`%k z|6Gvy5Li4Y6^g8Ys{wd&XP2H*(pK`7d{?Ji1hQ`wTq0UMIMi8#cmgWnIdLXQc{AlhGPd% zP?C`%ToX}E*wXQAg0*g61Qw0G#KgoH)|_#?c>Il~v&OwryesUBM^TeiUaw$oeD~pl zO#;E9r?mm&5fOEE4ITZOhFF0Ixi#7s4Kc7Is`L4CW)>C}R@Or4SNsBfHXmN0hz?A@ zdev*@hw+9bWY6T}My;f`4ZN#Ig zMDmSORCSR-!B0$_9BG>vA3qMJXt6Zqt%I3=(zE^@e#w{o0|J;C7|7~u*gbd#+a7CX zY8bR+h0yV$zi02>m)hY>^ry<^*Qgi;a+$8C7!=NBHzO3+p71W)BdxQMVg{H?HcGXu z0l8)Hcrft1W$k;%4U8An25xp7^^-U7-lKD7kikSW!h!F@hYx%{2mhr8ZBKtfRz`-W zukY8fF;n(u*NsSSE7!}I$zwqw@3a5Mf@b&-#>d7${2_T~IuT`?x|Wys@uNpL zOe+zc0IolihZKv>gWMV^?~;(dy|$*wJX5IWfa1exP*k$Qt4*f6~?vlr%s)M zG7QVjAJQ3H7@rwU$oOq+x3cH$8GlxZ@>LAEL+_%BiprmQ@ZaIoBo!6=L50oj6gl{0 z$*OA;iACdY>ZcrCN|Ca2w5_vqVPOFxtemXuXtAGa!6F;QH2|%7EEl*;;lj(`AIHX$ z67&W4mwmEuWmDCOi}|_&C`^g*d4B*zsFuj^37?Q}yEXqEFZO?*Ewf=mQetA_pAI|e zGj3;&@^uB|09KEG`O@Lt`S+nGZyu&nP1MrS8G8DZVscyNNv7i!5ZS8$^W6aL{~n)v zBt&6fnwpw&IB?*ntPrch6m_}QpGuOlm(nQUww9Q6st1dffP1`!Wm-ib1ZIw@*H%J2F4efY!)N_!<>GjfT(%T`Dx3JNJ3k5O$b zJM(~)OMu~?Eh{8+BnPfQC7b5beOU5oBCcuTq0=)7+DB8L}~=HtETF5D>s>4NLl zu?EwpNBe_8;UEm~SaxpSOuWZQz{avBy)2_T89fAp2USs~y05O^OwN;!&&#%^I6Fbw z71cE~Eac!qTOXTg--w>1WPx2b4yueKhx|kFoAAA-Xd6=hXi-sI`^rC*H270iEVths zJvB%uo}L>B8RB=D~h}u^KkqHI1L$3j^5L%A!PC4mAHc z893yiU3uaS;6r4pVT>83d)ZO$}t$@-e^$)wvCW^FPQ&jzMa|9bZ98BBmN z`%ay2N*?54n6i^)%#@+cG-kR-sDAWaznp59|6;3lVIzwo<^Ex-d{Zq=O+u)Pf?>4C zhugwj)FGP~H;NO&H!BkVCK29SZ(9i~m8YdP{`&cob)EJcM!I(3AAN^etGHqLqXJbATEyX~ z${lazcl(zRT>H-P{yEffNoK7)^*F z9R%E}P6=;+p5@ja98?E~sSh7|fxa77zW74Fw?hxkMGQ}X{6QHI1NHWCYwJt{waELbTr5U*e~yatzl48FVQM`7WpIy!L;HWfu_9Lg zgvnQtr_Y~b(;o{AJOkc}fk9DC&HLq@Ayic;MTtZr1-o)`8kC+?YNuF7DM#}ry^?;z z7iy!+mo9a`ZitDAp&|p((dr{ZiqlRawyL+c{O3nCKoGJ{7X0uUt_w{o^21FdkOa>a zdC$}z`L|u-d5D2Me@$ul_zbQD_if2t$9 zvoe&QCz23h|Co>d`?LHo;Qh@!^{n6=mH+!mVS=rMp+k~{rSC8F(0z9VLkIo{(D*&J zEvKn!pC?s4$&;Ya^nSs9=T z);Az3wqg9vHq3J|Xac+R7+3d97H8wVqBQ)83fZjqMNu9c-MD96V0z8}b9m^`S5klmOyE4}s_v?C(G7Ax&Ace}(2=TU$yY z36PAhMX3+W&6zr6oYl^t@_1Cov0t8(a<}72@Jk?G`{8t@A`hSZ0k6ZKFJ2-gPv`!)nvUx~|2>9( z`v3EPPh5)g2X~;bX_F;5)1)NB-AIDR6aOE7L+Ofojw ze0G<6XlSUpxjB+E+~U>!B|EX?&Y zV(#+F3Qn-+M7brl<;^I+6dUt*`EiuM>x&3?G|s;*Gc)r9$P!#a$1f;|b%{i|EouWl zHO!BfEmLy1tOBjAyRi>mmh5{!KEBt{(Z~HqJcWxyb^|AT@aWOgDZ5%xU|H=0D**2! z&$pRTBb!hN+4{>Mq97AfSLf&FgHXlv6+W(G^n?SMMT=HgjJIamlhPNU6Tsk#O${79 zt3-(yB~`{OWHpiVL0kpV25aT_tGc!_zzT8Z8+dCFwR~T9@7|4eSkp@~0*VJVonksE zHdKo;$7qWgk5c13Snk|!RY>N_d4)+a+PlP*2BrySo0ym|IO91fsv7#wnqM0ktQZS%C?1qN2_yjA2uUck401cf>-vBM7@~Ze~U^ z;$8LaEcpR#X=(a!LhaeaM2z7vg>x`hy>6)f#LJEc!XsW0XM zk*M+XXksVyfW{V8(61>eVF?dK20{=_=2L10wU7-85JNWY!C{y`eet5|{(V6Zv^W!# zmku@##p(M17y$3yy?ebE>hRLSmtU~quB?2Gd?LjYTZ%{n&&xQR@F-s!KLSjzF1|kp zX7tBkQ$i|E2E8|JYw7xjig@A-?z z)0GZ3b~LQ6eD$wSNTa@d1JnGjJ5$m@kIl&~a=HGG2)J4y<^^PnoB&vCNSXROMMSdO z+jrC+W?mf~9UYGN^#w$y4xG!GU%d$V6k2*pe6OKl^S#5*Asf);&cATA%kNKrh#3zY1hV)mP9moJZ_{~#J|mI-DY zj{yu|qIdvpAFZ2!{!L;N%kb!pEaGM_QaUFi%OwS;w}so+w)(9s1+`vHO^s?r+u9Cz ziG=YJhmZ#_G8V*MZybh2?CjpwaJQhi7>(eZFB3F7DP6m(vzHsfS66;WTV73kPqyf$ zrKAQmwcH=DO(mT>7dba!6(I0~pAtfhpE^4U3JR8e#TC6&d9;;u^z?e+i)94`t#98( zz}}4C#LCJVw_ZjJFj+&Rs;*9m;O*n1i$g7BFX{{`2R=`nDRD{-)YyUD=lU5cm&++R z{TFz~7wJkw85TWkB-qQ*h)Vt5bsfX>r%z1mAUx_${Mcl1(R3YEhx2i?f$%k|oN4Nj zVXi*&ya`5N)C9QyfsSy%))tnZ3fw+Jw*YjgNp>OaCk<7o6daYKbMI$o+$ykXvp*N8 z*GawA?|P9NCIKj4eT`iG#jd+JBJtd@HyQ8aCEGoL^_kZ8I2FMfNvhF*oF;B4aGbj_uot`w%sIdV2$bs8GtG z;3H$U`GcUZY*KVeF|IC(&W(hP8*hQuI7FXCJdJXZP1?_rp!2u)0B(7P>LJu~*upLV znP=vZBH%51-0_8aNP@nfm}3Bpjt9a3y4)I`tH$ z!L2#pcSg*nIW50iO98_GoNw%{v8mnvWCsf~Kvo=1NQ5E5!RkOiHkfoYi5&|P9yUAY zcX&we;fFe!nj!GbFJe`!yl%3QVeN*YLy{oyS@bZi-SV@yWp~PE&)k%#ul9!Q@vFirXf_NgV z6LIzOn4Js%^WI8aU8UMUm4a!=m_H;9cXlkeE*o$!8&4P{w4PDniu=V zTjD9W=PYIr2qZsgqfJ0W!3ZddiCT$rOAMwy!F2u^vkplnq7q5@HP4YKqp3#Llrvp7 z(bnN6ww(ZRL{{Vz;9=WE`Uc}C#b*u;q)frkL>9=9aJ2CovN7@~LryB^E8r!fF9AmO z0U+(buWwI6k{ufx6BZWM_w!o@54_2-&TQwi*GLT?WOr|JS_oMxXUgIHfP;WoI zPm1~(rwgJ4=ciak?lQL38KXw;0VZE@#H`Gl)?qseFxkVb_WJe2bNrMT^(R69Oesk# zvU4W=n4LvQu^+_jt{~h7Ju-q|pCQLi46FSDqv$0t*L)ywX(x5V<``O+SAGn`LxO=a z&|rG!6|LAXW%iSwjQmlHSqUWVj&g3GkB+ON0L?c^qp?z zQoWj$b#`XLzpI?FN#Z@uhH-FKE4l4X;p_44w9Hf;J-uwH0|a}3y;C(6>pf2!C8@jB zu3?o%4r28QZ?@5FJQiJcu43|%|l{+E`n z#&G&-Bjw^ob?|{$hEH`lZc=m;wXK?2bIN+T+H(%XQ_cb!EdMFQD4$7bH%2C67s;(q`h^xg@IENMDh|ghDR2CGZW_nHiko zAuNlUynk>7#DXnfLLF8LLKu22Ao-v|>wmginyLfm{wjo6+)$^6iW@y-LVGG?KlU|$ z!%VUvyoh~vcBEZ5Owp>))z#Iw1Ve*Re|)F__U@m`y&k=ceEkp5)R8BeL}>JSTgUHy z*QJH7B*v3;KlEql-YunG$Z$}ykXS(^qhp*vq#llq4IwKYi#k)~GnDjALtfqjsbmyA z2Qbk%VMGz!0-%?bfDeE(8A){f(o#}r+@$_!B_n<+F_Ey~%$?L&3$F!W#~Vz!_rSsB z`)JQ*gtRO4KCoLTQpg@a_!In?Y=J^MG#yhZ@dW)Ps4<0gen*}-vb4Le;_+x?kw}KB z+S$Sz#PUu%DTD1M9G~_W7kMh0(9mz_)u!scc_E3q$l)~?ee{lJqA`0;?_dqM5-Z4Y zP~YUJ$LWylsWm$jpL%bxnOqa`A|dU|+oSbI1H$+Af4_GB#am`@UVTGD7MId)6eX6g zN%VpmcDH9v+2z&K^2w>G0^X^wUyCn)NvToKhh&KM%c`oiDq}P>QFkK#3C5WB`LL54q~(WxWh z0g7Jq+5uaDzgs`ncnviYunO9jel0F~y{S%piC7E36l)^H?&0ZanDMgY+qduESs576 zIJJV21b7Vz#Ihys`_X7pOGruG z!1glp@h`8dyLt(Ly*HF6*!`}Dx}sv^(}IG6k!S9A?%F*wU^3uJO^2j~^Ja2L%iOyA z%ndGK;m%wrIlu_U{3qBL8ACtg_O+1rtnBRPBKM_C9MYQZ)DmqyV{%932v{n#IDGi> z#Uc~jfTBj3-}{enV+nnnOFuDp02T_`&qz8Rij!6D;#}@xu^jKTlrH-@mxE#11QigH z(YbTSgM&@kF~s6Es=tpPJ~+RC@1c-t`xeh%Po=Y+*8!mhKH=~DOJ!iEj*YopZF=#7 zT|Rg1Mv9A$Qfzj}BpQ2o$PzI<=>z~$x)QZT4Coe$p~gv?dZT6-=}`VW)Jiip!*?Si z$nqtWp5(PROG_J2d8u4Z+;Q;lGUio6M3SyY&+Sv`{6}LPg<`Kv_m}RkNOEQKczn3a zw&0SShEr+E8yRKw<@e|G-aTngc=pV^J>5K&#FN}kQwe;|fBn`iWV%y#+icXyPQRJN zackAOo}j?Ma4^p4dmqTj@S}BJmKWBm#WbZ8b6Ew&o2f1QET{6%>7ww+cWf;K@dDy_ z(j_W6z!rAn3k5Ktm@P2>6xvTMAIo95McOn6mmJJPbwP*vXa#Q6+|-DrG)!crwC)MrX8W8j@Y8HVaL zZV~Z|q6gWM?)4C)Yy98SMHpt*)@qj@0=k=9FXxX#!y)+lshv7)hL~B7bg6%!q`|h` z_vA5a^~p&u7uWM8GHS10SaH=B7bhg8JY%mYzd-j0K5L#c%|5=ZB>@wAdX&?<*BU#lG^1CSA!Fg>wC0MQs;cq*-voBm| zovlCglncIMYg zCaX5<5ialUujg4A%Q*}gzFi%NFitEM=NpiJsD#Exq^O|a;Clr~FFpfhR4t~VRr`$h zw?2BbhKHxj?|w#cVWEE@WQ;{4eSHGcC@wkB=l%lpp~>iKp^~Z3EfnB9|AYxwH+=@I z6eojGkl{!~IPl?lxyoW$Yau_O&Lrsx~O#iWUvF zq*(~ozZIe$AH8++W;zSx7V+KSWxTs}+NiVL*l9&L`A%Qauh`0(#0o-1MWNY6Yb%c7C1t5F+zSeSs9IZVYr~(3n zwq%O6_sK%}NQ=1aWtJ1iVVF4Hx*%bAhGN8(d*-gRp$^p&WZBeI&Px%z{%}jOJir$3 z%L7whTzn)XBz@ZMNqi%#+={W{PSpY(-Y86efEUp~Pj_`WV5<50cC@&FrY9@% z)ej#E>zQT*1O)h#>*}^tcB~f~6L42P?^Ved=;9()Tj%vkd8~c3SW;R#zquSEmDX>C z?wlB%--|*YwOD%;j`>O{ZAM^eNl8I8a_8+kchspZR2B~Q#lB-T$oXkR{)ifjKT3VP zMfg(}qOrx&fqPmpf4r7_+Xdtb9g)xFJ&HqvFeh zHhcF#15e!$zu;m*61 zoXi#;1`jasL#r0h7v`E_MnVPdQ=Xzdt=RDVSa2|N8i#=+9)Rj{qwVmX!R> zh-<>*1&lSJK4fa@|SM68-I}_e7#hhqHguWYGOd<*rHm=+reco0*$i z#+;0HZZVEikm~@{!J^{+Jy2Z>^Yewp#rK01_LsMzs~?^&B&ZO@hXy|o7QrQ8;!#a) zZ5}Ux%f~g0TJSV5>kL`Be0MiOXtU9Tp)V5jgNveMAdq}V0O25;|fSQtEdNU~ytOy)RP zQCWHZhfNxMSV;7Et9mzfXKcrwmkS^0TmD-57D=#INlB_bZQYXG>Fh?BgC37Sej-s+pERST2Jh zF|tr`H~MkVedm8OF(IQ&ilsD_eQvzz0ge!dV8>Pe&)Z_n)RX<8#?>W`-n?}S$E4#N zkB5FNS{0FSPC`;Pdk!X+lekb8*9V~t zUDrQwyunXc3ltR=TB%vHaxpFQEJcg&Q8iOk1_L#gs%Ab zcsKYe+djzj5T%PY6Gxv(g5W(aVt>x5CTxB~oKys6>|DNb<>^&fIXO!ASzB8*;8uFj z9@PMn_4#yF35M70G?v%$HC4HfK}e9<%3x5{e$3{{?Bvw4NL33 zeqkoRS=v$gnL4w^t(q@)UQ{lWd>vD28HlbS^S`Q?Kk2Hoya>`pb!49X^ODDRld9WP z$x+_prDaJ;WVPUx%6MbMAV-25_c_ZqJD;37aLsO3$J2%F3OC=0u!9^ePR5Eyl)ouU zeP64??b7`xsWB+&r06#|pmu&}kIdYOOB>qj%-Z5<_A>yX_YfK9Y+!s0I$9&kH?hk9 zCmMnvlL^gir-5f+-?{iOCn(`HRyxz>);(9NOGv5B&WuAwgGg33zLmLmQdA3qN@} z{n&ncd&sBYzpt%0aZz^~#6I16LmDMeCmU`#U;!@KDMV{JqrKzhBglVRMQ; z^Pm6cf`|Y6ms5D~KR*?_?(b5H#g-BOu23vTu;`x`7|YSj_V?Rx7!E}J-~6zcRqDB- Wj`D4)rLPG1v(3cP_>Q4-^#1~!;+dNO literal 0 HcmV?d00001 diff --git a/docs/search_index.json b/docs/search_index.json index eeee6fd..7114284 100644 --- a/docs/search_index.json +++ b/docs/search_index.json @@ -1 +1 @@ -[["index.html", "Introduction to Reproducible Research About this Course Learning Objectives 0.1 Available course formats", " Introduction to Reproducible Research April, 2023 About this Course This course will introduce the learner to the basic concepts of reproducibility and replication of scientific research. The premise of the course is that reproducibility failure can be traced to many different root causes and being able to identify these root causes will provide the basis for preventing future failure. The course will provide case studies from three historical data analyses depicting what constitutes a failure of reproducibility. The motivating case studies come from (1) the Harvard Six Cities Study of air pollution and mortality; (2) the National Morbidity and Mortality Air Pollution Study, and (3) a study describing the development of microarray-based genomic signatures for predicting chemotherapy response. These case studies highlight various aspects of data analysis that can lead to reproducibility failure, such as workflow management, data analytic modeling choices, stakeholder engagement, team organization, and communication. Learning Objectives Upon completing this module, a learner will be able to Describe the requirements for reproducible analyses Identify stakeholders affected by reproducibility of an analysis (i.e. collaborators, students, other researchers) Describe the impact of reproducibility on various data analysis stakeholders Characterize the difference between reproducibility and replication Identify organizational or analytic problems that can lead to non-reproducible research 0.1 Available course formats This course is available in multiple formats which allows you to take it in the way that best suites your needs. You can take it for certificate which can be for free or fee. The material for this course can be viewed without login requirement on this Bookdown website. This format might be most appropriate for you if you rely on screen-reader technology. This course can be taken for free certification through Leanpub. This course can be taken on Coursera for certification here (but it is not available for free on Coursera). Our courses are open source, you can find the source material for this course on GitHub. "],["introduction.html", "Chapter 1 Introduction 1.1 Motivation 1.2 Target Audience 1.3 Curriculum", " Chapter 1 Introduction 1.1 Motivation 1.2 Target Audience The course is intended for … 1.3 Curriculum The course covers… devtools::session_info() ## ─ Session info ─────────────────────────────────────────────────────────────── ## setting value ## version R version 4.0.2 (2020-06-22) ## os Ubuntu 20.04.5 LTS ## system x86_64, linux-gnu ## ui X11 ## language (EN) ## collate en_US.UTF-8 ## ctype en_US.UTF-8 ## tz Etc/UTC ## date 2023-04-11 ## ## ─ Packages ─────────────────────────────────────────────────────────────────── ## package * version date lib source ## assertthat 0.2.1 2019-03-21 [1] RSPM (R 4.0.5) ## bookdown * 0.24 2023-03-28 [1] Github (rstudio/bookdown@88bc4ea) ## bslib 0.4.2 2022-12-16 [1] CRAN (R 4.0.2) ## cachem 1.0.7 2023-02-24 [1] CRAN (R 4.0.2) ## callr 3.5.0 2020-10-08 [1] RSPM (R 4.0.2) ## cli 3.6.1 2023-03-23 [1] CRAN (R 4.0.2) ## crayon 1.3.4 2017-09-16 [1] RSPM (R 4.0.0) ## desc 1.2.0 2018-05-01 [1] RSPM (R 4.0.3) ## devtools 2.3.2 2020-09-18 [1] RSPM (R 4.0.3) ## digest 0.6.25 2020-02-23 [1] RSPM (R 4.0.0) ## ellipsis 0.3.1 2020-05-15 [1] RSPM (R 4.0.3) ## evaluate 0.20 2023-01-17 [1] CRAN (R 4.0.2) ## fansi 0.4.1 2020-01-08 [1] RSPM (R 4.0.0) ## fastmap 1.1.1 2023-02-24 [1] CRAN (R 4.0.2) ## fs 1.5.0 2020-07-31 [1] RSPM (R 4.0.3) ## glue 1.4.2 2020-08-27 [1] RSPM (R 4.0.5) ## hms 0.5.3 2020-01-08 [1] RSPM (R 4.0.0) ## htmltools 0.5.5 2023-03-23 [1] CRAN (R 4.0.2) ## jquerylib 0.1.4 2021-04-26 [1] CRAN (R 4.0.2) ## jsonlite 1.7.1 2020-09-07 [1] RSPM (R 4.0.2) ## knitr 1.33 2023-03-28 [1] Github (yihui/knitr@a1052d1) ## lifecycle 1.0.3 2022-10-07 [1] CRAN (R 4.0.2) ## magrittr 2.0.3 2022-03-30 [1] CRAN (R 4.0.2) ## memoise 2.0.1 2021-11-26 [1] CRAN (R 4.0.2) ## ottrpal 1.0.1 2023-03-28 [1] Github (jhudsl/ottrpal@151e412) ## pillar 1.9.0 2023-03-22 [1] CRAN (R 4.0.2) ## pkgbuild 1.1.0 2020-07-13 [1] RSPM (R 4.0.2) ## pkgconfig 2.0.3 2019-09-22 [1] RSPM (R 4.0.3) ## pkgload 1.1.0 2020-05-29 [1] RSPM (R 4.0.3) ## prettyunits 1.1.1 2020-01-24 [1] RSPM (R 4.0.3) ## processx 3.4.4 2020-09-03 [1] RSPM (R 4.0.2) ## ps 1.4.0 2020-10-07 [1] RSPM (R 4.0.2) ## R6 2.4.1 2019-11-12 [1] RSPM (R 4.0.0) ## readr 1.4.0 2020-10-05 [1] RSPM (R 4.0.2) ## remotes 2.2.0 2020-07-21 [1] RSPM (R 4.0.3) ## rlang 1.1.0 2023-03-14 [1] CRAN (R 4.0.2) ## rmarkdown 2.10 2023-03-28 [1] Github (rstudio/rmarkdown@02d3c25) ## rprojroot 2.0.3 2022-04-02 [1] CRAN (R 4.0.2) ## sass 0.4.5 2023-01-24 [1] CRAN (R 4.0.2) ## sessioninfo 1.1.1 2018-11-05 [1] RSPM (R 4.0.3) ## stringi 1.5.3 2020-09-09 [1] RSPM (R 4.0.3) ## stringr 1.4.0 2019-02-10 [1] RSPM (R 4.0.3) ## testthat 3.0.1 2023-03-28 [1] Github (R-lib/testthat@e99155a) ## tibble 3.2.1 2023-03-20 [1] CRAN (R 4.0.2) ## usethis 1.6.3 2020-09-17 [1] RSPM (R 4.0.2) ## utf8 1.1.4 2018-05-24 [1] RSPM (R 4.0.3) ## vctrs 0.6.1 2023-03-22 [1] CRAN (R 4.0.2) ## withr 2.3.0 2020-09-22 [1] RSPM (R 4.0.2) ## xfun 0.26 2023-03-28 [1] Github (yihui/xfun@74c2a66) ## yaml 2.2.1 2020-02-01 [1] RSPM (R 4.0.3) ## ## [1] /usr/local/lib/R/site-library ## [2] /usr/local/lib/R/library "],["a-new-chapter.html", "Chapter 2 A new chapter 2.1 Learning Objectives 2.2 Libraries 2.3 Topic of Section 2.4 Code examples 2.5 Image example 2.6 Video examples 2.7 File examples 2.8 Website Examples 2.9 Citation examples 2.10 Stylized boxes 2.11 Dropdown summaries 2.12 Print out session info", " Chapter 2 A new chapter If you haven’t yet read the getting started Wiki pages; start there. To see the rendered version of this chapter and the rest of the template, see here: https://jhudatascience.org/OTTR_Template/. Every chapter needs to start out with this chunk of code: 2.1 Learning Objectives Every chapter also needs Learning objectives that will look like this: This chapter will cover: {You can use https://tips.uark.edu/using-blooms-taxonomy/ to define some learning objectives here} {Another learning objective} 2.2 Libraries For this chapter, we’ll need the following packages attached: *Remember to add any additional packages you need to your course’s own docker image. library(magrittr) 2.3 Topic of Section You can write all your text in sections like this, using ## to indicate a new header. you can use additional pound symbols to create lower levels of headers. See here for additional general information about how you can format text within R Markdown files. In addition, see here for more in depth and advanced options. 2.3.1 Subtopic Here’s a subheading (using three pound symbols) and some text in this subsection! 2.4 Code examples You can demonstrate code like this: output_dir <- file.path("resources", "code_output") if (!dir.exists(output_dir)) { dir.create(output_dir) } And make plots too: hist_plot <- hist(iris$Sepal.Length) You can also save these plots to file: png(file.path(output_dir, "test_plot.png")) hist_plot ## $breaks ## [1] 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 ## ## $counts ## [1] 5 27 27 30 31 18 6 6 ## ## $density ## [1] 0.06666667 0.36000000 0.36000000 0.40000000 0.41333333 0.24000000 0.08000000 ## [8] 0.08000000 ## ## $mids ## [1] 4.25 4.75 5.25 5.75 6.25 6.75 7.25 7.75 ## ## $xname ## [1] "iris$Sepal.Length" ## ## $equidist ## [1] TRUE ## ## attr(,"class") ## [1] "histogram" dev.off() ## png ## 2 2.5 Image example How to include a Google slide. It’s simplest to use the ottrpal package: But if you have the slide or some other image locally downloaded you can also use HTML like this: 2.6 Video examples You may also want to embed videos in your course. If alternatively, you just want to include a link you can do so like this: Check out this link to a video using markdown syntax. 2.6.1 Using knitr To embed videos in your course, you can use knitr::include_url() like this: Note that you should use echo=FALSE in the code chunk because we don’t want the code part of this to show up. If you are unfamiliar with how R Markdown code chunks work, read this. 2.6.2 Using HTML 2.7 File examples You can again use simple markdown syntax to just include a link to a file like so: A file. Alternatively you can embed files like PDFs. 2.7.1 Using knitr 2.7.2 Using HTML 2.8 Website Examples Yet again you can use a link to a website like so: A Website You might want to have users open a website in a new tab by default, especially if they need to reference both the course and a resource at once. A Website Or, you can embed some websites. 2.8.1 Using knitr This works: 2.8.2 Using HTML If you’d like the URL to show up in a new tab you can do this: <a href="https://www.linkedin.com" target="_blank">LinkedIn</a> 2.9 Citation examples We can put citations at the end of a sentence like this (Allaire et al. 2021). Or multiple citations Xie, Allaire, and Grolemund (2018). but they need a ; separator (Allaire et al. 2021; Xie, Allaire, and Grolemund 2018). In text, we can put citations like this Allaire et al. (2021). 2.10 Stylized boxes Occasionally, you might find it useful to emphasize a particular piece of information. To help you do so, we have provided css code and images (no need for you to worry about that!) to create the following stylized boxes. You can use these boxes in your course with either of two options: using HTML code or Pandoc syntax. 2.10.1 Using rmarkdown container syntax The rmarkdown package allows for a different syntax to be converted to the HTML that you just saw and also allows for conversion to LaTeX. See the Bookdown documentation for more information (Xie, Dervieux, and Riederer 2020). Note that Bookdown uses Pandoc. ::: {.notice} Note using rmarkdown syntax. ::: Note using rmarkdown syntax. As an example you might do something like this: Please click on the subsection headers in the left hand navigation bar (e.g., 2.1, 4.3) a second time to expand the table of contents and enable the scroll_highlight feature (see more) 2.10.2 Using HTML To add a warning box like the following use: <div class = "notice"> Followed by the text you want inside </div> This will create the following: Followed by the text you want inside Here is a <div class = \"warning\"> box: Note text Here is a <div class = \"github\"> box: GitHub text Here is a <div class = \"dictionary\"> box: dictionary text Here is a <div class = \"reflection\"> box: reflection text 2.11 Dropdown summaries You can hide additional information in a dropdown menu Here’s more words that are hidden. 2.12 Print out session info You should print out session info when you have code for reproducibility purposes. devtools::session_info() ## ─ Session info ─────────────────────────────────────────────────────────────── ## setting value ## version R version 4.0.2 (2020-06-22) ## os Ubuntu 20.04.5 LTS ## system x86_64, linux-gnu ## ui X11 ## language (EN) ## collate en_US.UTF-8 ## ctype en_US.UTF-8 ## tz Etc/UTC ## date 2023-04-11 ## ## ─ Packages ─────────────────────────────────────────────────────────────────── ## package * version date lib source ## assertthat 0.2.1 2019-03-21 [1] RSPM (R 4.0.5) ## bookdown * 0.24 2023-03-28 [1] Github (rstudio/bookdown@88bc4ea) ## bslib 0.4.2 2022-12-16 [1] CRAN (R 4.0.2) ## cachem 1.0.7 2023-02-24 [1] CRAN (R 4.0.2) ## callr 3.5.0 2020-10-08 [1] RSPM (R 4.0.2) ## cli 3.6.1 2023-03-23 [1] CRAN (R 4.0.2) ## crayon 1.3.4 2017-09-16 [1] RSPM (R 4.0.0) ## curl 4.3 2019-12-02 [1] RSPM (R 4.0.3) ## desc 1.2.0 2018-05-01 [1] RSPM (R 4.0.3) ## devtools 2.3.2 2020-09-18 [1] RSPM (R 4.0.3) ## digest 0.6.25 2020-02-23 [1] RSPM (R 4.0.0) ## ellipsis 0.3.1 2020-05-15 [1] RSPM (R 4.0.3) ## evaluate 0.20 2023-01-17 [1] CRAN (R 4.0.2) ## fansi 0.4.1 2020-01-08 [1] RSPM (R 4.0.0) ## fastmap 1.1.1 2023-02-24 [1] CRAN (R 4.0.2) ## fs 1.5.0 2020-07-31 [1] RSPM (R 4.0.3) ## glue 1.4.2 2020-08-27 [1] RSPM (R 4.0.5) ## highr 0.8 2019-03-20 [1] RSPM (R 4.0.3) ## hms 0.5.3 2020-01-08 [1] RSPM (R 4.0.0) ## htmltools 0.5.5 2023-03-23 [1] CRAN (R 4.0.2) ## httr 1.4.2 2020-07-20 [1] RSPM (R 4.0.3) ## jquerylib 0.1.4 2021-04-26 [1] CRAN (R 4.0.2) ## jsonlite 1.7.1 2020-09-07 [1] RSPM (R 4.0.2) ## knitr 1.33 2023-03-28 [1] Github (yihui/knitr@a1052d1) ## lifecycle 1.0.3 2022-10-07 [1] CRAN (R 4.0.2) ## magrittr * 2.0.3 2022-03-30 [1] CRAN (R 4.0.2) ## memoise 2.0.1 2021-11-26 [1] CRAN (R 4.0.2) ## ottrpal 1.0.1 2023-03-28 [1] Github (jhudsl/ottrpal@151e412) ## pillar 1.9.0 2023-03-22 [1] CRAN (R 4.0.2) ## pkgbuild 1.1.0 2020-07-13 [1] RSPM (R 4.0.2) ## pkgconfig 2.0.3 2019-09-22 [1] RSPM (R 4.0.3) ## pkgload 1.1.0 2020-05-29 [1] RSPM (R 4.0.3) ## prettyunits 1.1.1 2020-01-24 [1] RSPM (R 4.0.3) ## processx 3.4.4 2020-09-03 [1] RSPM (R 4.0.2) ## ps 1.4.0 2020-10-07 [1] RSPM (R 4.0.2) ## R6 2.4.1 2019-11-12 [1] RSPM (R 4.0.0) ## readr 1.4.0 2020-10-05 [1] RSPM (R 4.0.2) ## remotes 2.2.0 2020-07-21 [1] RSPM (R 4.0.3) ## rlang 1.1.0 2023-03-14 [1] CRAN (R 4.0.2) ## rmarkdown 2.10 2023-03-28 [1] Github (rstudio/rmarkdown@02d3c25) ## rprojroot 2.0.3 2022-04-02 [1] CRAN (R 4.0.2) ## sass 0.4.5 2023-01-24 [1] CRAN (R 4.0.2) ## sessioninfo 1.1.1 2018-11-05 [1] RSPM (R 4.0.3) ## stringi 1.5.3 2020-09-09 [1] RSPM (R 4.0.3) ## stringr 1.4.0 2019-02-10 [1] RSPM (R 4.0.3) ## testthat 3.0.1 2023-03-28 [1] Github (R-lib/testthat@e99155a) ## tibble 3.2.1 2023-03-20 [1] CRAN (R 4.0.2) ## usethis 1.6.3 2020-09-17 [1] RSPM (R 4.0.2) ## utf8 1.1.4 2018-05-24 [1] RSPM (R 4.0.3) ## vctrs 0.6.1 2023-03-22 [1] CRAN (R 4.0.2) ## withr 2.3.0 2020-09-22 [1] RSPM (R 4.0.2) ## xfun 0.26 2023-03-28 [1] Github (yihui/xfun@74c2a66) ## yaml 2.2.1 2020-02-01 [1] RSPM (R 4.0.3) ## ## [1] /usr/local/lib/R/site-library ## [2] /usr/local/lib/R/library "],["about-the-authors.html", "About the Authors", " About the Authors These credits are based on our course contributors table guidelines.     Credits Names Pedagogy Lead Content Instructor(s) Roger Peng Technical Template Publishing Engineers Candace Savonen, Carrie Wright Publishing Maintenance Engineer Candace Savonen Technical Publishing Stylists Carrie Wright, Candace Savonen Package Developers (ottrpal) Candace Savonen, John Muschelli, Carrie Wright Art and Design Illustrator(s) Candace Savonen, Carrie Wright Funding Funder(s) NIH Grant R25GM141505   ## ─ Session info ─────────────────────────────────────────────────────────────── ## setting value ## version R version 4.0.2 (2020-06-22) ## os Ubuntu 20.04.5 LTS ## system x86_64, linux-gnu ## ui X11 ## language (EN) ## collate en_US.UTF-8 ## ctype en_US.UTF-8 ## tz Etc/UTC ## date 2023-04-11 ## ## ─ Packages ─────────────────────────────────────────────────────────────────── ## package * version date lib source ## assertthat 0.2.1 2019-03-21 [1] RSPM (R 4.0.5) ## bookdown * 0.24 2023-03-28 [1] Github (rstudio/bookdown@88bc4ea) ## bslib 0.4.2 2022-12-16 [1] CRAN (R 4.0.2) ## cachem 1.0.7 2023-02-24 [1] CRAN (R 4.0.2) ## callr 3.5.0 2020-10-08 [1] RSPM (R 4.0.2) ## cli 3.6.1 2023-03-23 [1] CRAN (R 4.0.2) ## crayon 1.3.4 2017-09-16 [1] RSPM (R 4.0.0) ## desc 1.2.0 2018-05-01 [1] RSPM (R 4.0.3) ## devtools 2.3.2 2020-09-18 [1] RSPM (R 4.0.3) ## digest 0.6.25 2020-02-23 [1] RSPM (R 4.0.0) ## ellipsis 0.3.1 2020-05-15 [1] RSPM (R 4.0.3) ## evaluate 0.20 2023-01-17 [1] CRAN (R 4.0.2) ## fastmap 1.1.1 2023-02-24 [1] CRAN (R 4.0.2) ## fs 1.5.0 2020-07-31 [1] RSPM (R 4.0.3) ## glue 1.4.2 2020-08-27 [1] RSPM (R 4.0.5) ## htmltools 0.5.5 2023-03-23 [1] CRAN (R 4.0.2) ## jquerylib 0.1.4 2021-04-26 [1] CRAN (R 4.0.2) ## jsonlite 1.7.1 2020-09-07 [1] RSPM (R 4.0.2) ## knitr 1.33 2023-03-28 [1] Github (yihui/knitr@a1052d1) ## magrittr 2.0.3 2022-03-30 [1] CRAN (R 4.0.2) ## memoise 2.0.1 2021-11-26 [1] CRAN (R 4.0.2) ## pkgbuild 1.1.0 2020-07-13 [1] RSPM (R 4.0.2) ## pkgload 1.1.0 2020-05-29 [1] RSPM (R 4.0.3) ## prettyunits 1.1.1 2020-01-24 [1] RSPM (R 4.0.3) ## processx 3.4.4 2020-09-03 [1] RSPM (R 4.0.2) ## ps 1.4.0 2020-10-07 [1] RSPM (R 4.0.2) ## R6 2.4.1 2019-11-12 [1] RSPM (R 4.0.0) ## remotes 2.2.0 2020-07-21 [1] RSPM (R 4.0.3) ## rlang 1.1.0 2023-03-14 [1] CRAN (R 4.0.2) ## rmarkdown 2.10 2023-03-28 [1] Github (rstudio/rmarkdown@02d3c25) ## rprojroot 2.0.3 2022-04-02 [1] CRAN (R 4.0.2) ## sass 0.4.5 2023-01-24 [1] CRAN (R 4.0.2) ## sessioninfo 1.1.1 2018-11-05 [1] RSPM (R 4.0.3) ## stringi 1.5.3 2020-09-09 [1] RSPM (R 4.0.3) ## stringr 1.4.0 2019-02-10 [1] RSPM (R 4.0.3) ## testthat 3.0.1 2023-03-28 [1] Github (R-lib/testthat@e99155a) ## usethis 1.6.3 2020-09-17 [1] RSPM (R 4.0.2) ## withr 2.3.0 2020-09-22 [1] RSPM (R 4.0.2) ## xfun 0.26 2023-03-28 [1] Github (yihui/xfun@74c2a66) ## yaml 2.2.1 2020-02-01 [1] RSPM (R 4.0.3) ## ## [1] /usr/local/lib/R/site-library ## [2] /usr/local/lib/R/library "],["references.html", "Chapter 3 References", " Chapter 3 References "],["404.html", "Page not found", " Page not found The page you requested cannot be found (perhaps it was moved or renamed). You may want to try searching to find the page's new location, or use the table of contents to find the page you are looking for. "]] +[["index.html", "Introduction to Reproducible Research About this Course Learning Objectives Available course formats", " Introduction to Reproducible Research April, 2023 About this Course This course will introduce the learner to the basic concepts of reproducibility and replication of scientific research. The premise of the course is that reproducibility failure can be traced to many different root causes and being able to identify these root causes will provide the basis for preventing future failure. The course will provide case studies from three historical data analyses depicting what constitutes a failure of reproducibility. The motivating case study comes from a study describing the development of microarray-based genomic signatures for predicting chemotherapy response. This case study highlights various aspects of data analysis that can lead to reproducibility failure, such as workflow management, data analytic modeling choices, stakeholder engagement, team organization, and communication. Learning Objectives Upon completing this module, a learner will be able to Describe the requirements for reproducible analyses Identify stakeholders affected by reproducibility of an analysis (i.e. collaborators, students, other researchers) Describe the impact of reproducibility on various data analysis stakeholders Characterize the difference between reproducibility and replication Identify organizational or analytic problems that can lead to non-reproducible research Available course formats This course is available in multiple formats which allows you to take it in the way that best suites your needs. You can take it for certificate which can be for free or fee. The material for this course can be viewed without login requirement on this Bookdown website. This format might be most appropriate for you if you rely on screen-reader technology. This course can be taken for free certification through Leanpub. This course can be taken on Coursera for certification here (but it is not available for free on Coursera). Our courses are open source, you can find the source material for this course on GitHub. "],["introduction.html", "Chapter 1 Introduction 1.1 What’s Wrong with Replication? 1.2 Reproducibility to the Rescue 1.3 Reproducibility Guidelines 1.4 Summary", " Chapter 1 Introduction This chapter will be about the concept of reproducible research and we will cover some basic concepts and ideas that are related to reproducible research. Before we get to reproducibility, we need to cover a little background with respect to how science works (even if you’re not a scientist, this is important). The basic idea is that in science, replication is the most important element of verifying and validating findings. So if you claim that X causes Y, or that Vitamin C improves disease, or that something causes a problem, what happens is that other scientists that are independent of you will try to investigate that same question and see if they come up with a similar result. If lots of different people come up with the same result and replicate the original finding, then we tend to think that the original finding was probably true and that this is a real relationship or real finding. The ultimate standard in strengthening scientific evidence is replication. The goal is to have independent people to do independent things with different data, different methods, and different laboratories and see if you get the same result. There’s a sense that if a relationship in nature is truly there, then it should be robust to having different people discover it in different ways. Replication is particularly important in areas where findings can have big policy impacts or can influence regulatory types of decisions. 1.1 What’s Wrong with Replication? So what’s wrong with replication? There’s really nothing wrong with it. This is what science has been doing for a long time, through hundreds of years. And there’s nothing wrong with it today. But the problem is that it’s becoming more and more challenging to do replication or to replicate other studies. Part of the reason is because studies are getting bigger and bigger. In order to do big studies you need a lot of money and so, well, there’s a lot of money involved! If you want to do ten versions of the same study, you need ten times as much money and there’s not as much money around as there used to be. Sometimes it’s difficult to replicate a study because if the original study took 20 years to do, it’s difficult to wait around another 20 years for replication. Some studies are just plain unique, such as studying the impact of a massive earthquake in a very specific location and time. If you’re looking at a unique situation in time or a unique population, you can’t readily replicate that situation. There are a lot of good reasons why you can’t replicate a study. If you can’t replicate a study, is the alternative just to do nothing, just let that study stand by itself? The idea behind a reproducible research is to create a kind of minimum standard or a middle ground where we won’t be replicating a study, but maybe we can do something in between. The basic problem is that you have the gold standard, which is replication, and then you have the worst standard which is doing nothing. What can we do that’s in between the gold standard and doing nothing? That is where reproducibility comes in. That’s how we can kind of bridge the gap between replication and nothing. In non-research settings, often full replication isn’t even the point. Often the goal is to preserve something to the point where anybody in an organization can repeat what you did (for example, after you leave the organization). In this case, reproducibility is key to maintaining the history of a project and making sure that every step along the way is clear. 1.2 Reproducibility to the Rescue Why do we need this kind of middle ground? We haven’t clearly defined reproducibility yet, but the basic idea is that you need to make the data available for the original study and the computational methods available so that other people can look at your data and run the kind of analysis that you’ve run, and come to the same findings that you found. What reproducible research is about is a validation of the data analysis. Because you’re not collecting independent data using independent methods, it’s a little bit more difficult to validate the scientific question itself. But if you can take someone’s data and reproduce their findings, then you can, in some sense, validate the data analysis. This involves having the data and the code because more likely than not, the analysis will have been done on the computer using some sort of programming language, like R. So you can take their code and their data and reproduce the findings that they come up with. Then you can at least have confidence that the analysis was done appropriately and that the correct methods were used. Recently, there’s been a lot of discussion of reproducibility in the media and in the scientific literature. The journal Science had a special issue on reproducibility and data replication. Other journals have updated policies on publication to encourage reproducibility. In 2012, a feature on the TV show 60 minutes looked at a major incident at Duke University where many results involving a promising cancer test were found to be not reproducible (more on that later). This led to a number of studies and clinical trials having to be stopped, followed by an investigation which is still ongoing. 1.3 Reproducibility Guidelines Finally, the Institute of Medicine (now the National Academy of Medicine), in response to a lot of events involving reproducibility of scientific studies, issued a report saying that best practices should be done to promote and encourage reproducibility, particularly in what’s called ’omics based research, such as genomics, proteomics, other similar areas involving high-throughput biological measurements. This was a very important report. Of the many recommendations that the IOM made, the key ones were that Data and metadata need to be made available; Computer code should be fully specified, so that people can examine it to see what was done; All the steps of the computational analysis, including any preprocessing of data, should be fully described so that people can study it and reproduce it. We will expand on these ideas in the chapters that follow. 1.4 Summary Replication, whereby scientific questions are examined and verified independently by different scientists, is the gold standard for scientific validity. Replication can be difficult and often there are no resources to independently replicate a study. Reproducibility, whereby data and code are re-analyzed by independent scientists to obtain the same results of the original investigator, is a reasonable minimum standard when replication is not possible. "],["from-x-to-computational-x.html", "Chapter 2 From “X” to “Computational X” 2.1 Example: Air Pollution and Health 2.2 The Data Science Pipeline 2.3 Elements of Reproducibility 2.4 Authors and Readers 2.5 Summary", " Chapter 2 From “X” to “Computational X” What is driving this need for a “reproducibility middle ground” between replication and doing nothing? For starters, there are a lot of new technologies on the scene and in many different fields of study including, biology, chemistry and environmental science. These technologies allow us to collect data at a much higher throughput so we end up with these very complex and very high dimensional data sets. These datasets can be collected almost instantaneously compared to even just ten years ago—the technology has allowed us to create huge data sets at essentially the touch of a button. Furthermore, we the computing power to take existing (already huge) databases and merge them into even bigger and bigger databases. Finally, the massive increase in computing power has allowed us to implement more sophisticated and complex analysis routines. The analyses themselves, the models that we fit and the algorithms that we run, are much much more complicated than they used to be. Having a basic understanding of these algorithms is difficult, even for a sophisticated person, and it’s almost impossible to describe these algorithms with words alone. Understanding what someone did in a data analysis now requires looking at code and scrutinizing the computer programs that people used. The bottom line with all these different trends is that for every field “X”, there is now “Computational X”. There’s computational biology, computational astronomy—whatever it is you want, there is a computational version of it. 2.1 Example: Air Pollution and Health One example of an area were reproducibility is important comes from the area of air pollution and health. Air pollution and health is a big field and it involves a confluence of features that emphasize the need for reproducibility. The first feature is that we are estimating very small, but very important, public health effects in the presence of a numerous much stronger signals. You can think about air pollution as something that’s perhaps harmful, but even if it were harmful there are likely many other things that are going to be more harmful that you have to worry about. Pollution is going to be at the very top of the list of things that are going to harm you. In other words, there’s an inherently weak signal there. Second, the results of a lot of air pollution research inform substantial policy decisions. Many federal air pollution regulations in the United States are based on scientific research in this area and these regulations can affect a lot of stakeholders in government and industry. Finally, we use a lot of complex statistical methods to do these studies and these statistical methods are subsequently subjected to intense scrutiny. The combination of an inherently weak signal, substantial policy impacts, and complex statistical methods almost require that the research that we do be reproducible. 2.2 The Data Science Pipeline The basic issue is when you read a description of a data analysis, such as in an article or a technical report, for the most part, what you get is the report and nothing else. Of course, everyone knows that behind the scenes there’s a lot that went into this report and that’s what I call the data science pipeline. In this pipeline, there are two “actors”: the author of the report/article and the reader. On the left side, the author is going from left to right along this pipeline. The reader is going from right to left. If you’re the reader you read the article and you want to know more about what happened: Where is the data? What was used here? The basic idea behind reproducibility is to focus on the analytic data and the computational results. With reproducibility the goal is to allow the author of a report and the reader of that report to “meet in the middle”. 2.3 Elements of Reproducibility What do we need for reproducibility? There’s a variety of ways to talk about this, but one basic definition that we’ve come up with is that there are four things that are required to make results reproducible: Analytic data. The data that were used for the analysis that was presented should be available for others to access. This is different from the raw data because very often in a data analysis the raw data are not all used for the analysis, but rather some subset is used. It may be interesting to see the raw data but impractical to actually have it. Analytic data is key to examining the data analysis. Analytic code. The analytic code is the code that was applied to the analytic data to produce the key results. This may be preprocessing code, regression modeling code, or really any other code used to produce the results from the analytic data. Documentation. Documentation of that code and the data is very important. Distribution. Finally, there needs to be some standard means of distribution, so all this data in the code is easily accessible. 2.4 Authors and Readers It is important to realize that there are multiple players when you talk about reproducibility–there are different types of parties that have different types of interests. There are authors who produce research and they want to make their research reproducible. There are also readers of research and they want to reproduce that work. Everyone needs tools to make their lives easier. One current challenge is that authors of research have to undergo considerable effort to make their results available to a wide audience. Publishing data and code today is not necessarily a trivial task. Although there are a number of resources available now, that were not available even five years ago, it’s still a bit of a challenge to get things out on the web (or at least distributed widely). Resources like GitHub and various data repositories have made a big difference, but there is still a ways to go with respect to building up the public reproducibility infrastructure. Furthermore, even when data and code are available, readers often have to download the data, download the code, and then they have to piece everything together, usually by hand. It’s not always an easy task to put the data and code together. Also, readers may not have the same computational resources that the original authors did. If the original authors used an enormous computing cluster, for example, to do their analysis, the readers may not have that same enormous computing cluster at their disposal. It may be difficult for readers to reproduce the same results. Generally the toolbox for doing reproducible research is small, although it’s definitely growing. In practice, authors often just throw things up on the web. There are journals and supplementary materials, but they are famously disorganized. There are only a few central databases that authors can take advantage of to post their data and make it available. So if you’re working in a field that has a central database that everyone uses, that’s great. If you’re not, then you have to assemble your own resources. We will discuss the tools for doing reproducible research in the third course in this series. 2.5 Summary The process of conducting and disseminating research can be depicted as a “data science pipeline” Readers and consumers of data science research are typically not privy to the details of the data science pipeline One view of reproducibility is that it gives research consumers partial access to the raw pipeline elements. "],["learning-from-failure.html", "Chapter 3 Learning from Failure", " Chapter 3 Learning from Failure Learning about how data analyses succeeds or fails (but more importantly, fails) is extremely challenging without actually going through the process yourself. I don’t think I ever learned about it except through first hand experience, which took place over the course of years. There are a few reasons for this that I have observed over time: Success in scientific data analysis is usually concerned with whether the claims made based on the results are true or not. If the results feel true, and the analysis appears rigorous, then that’s usually the end of the discussion. Focus is put on the result and what should come next. The underlying idea here is not necessarily misguided: Progress in science depends on independent replication, and any given analysis cannot be assigned too much weight. When analyses fail, the results are usually vague and confusing. Furthermore, the public rarely finds out about them because they are not published. This is mostly due to human nature: it’s difficult to motivate oneself to write about an experience that was inconclusive and perhaps incoherent. It can also be embarrassing if honest mistakes were made. Publication of negative studies is a separate matter, because a truly negative study is, in fact, conclusive. But often, we don’t even have that much clarity. In the rare cases where we do find out about data analysis failures, the focus is often on who or what is to blame. In cases where criminal activity has taken place, this is an important aspect. However, identifying who or what is to blame usually doesn’t provide us with generalizable knowledge that we can apply to our own data analyses. The underlying assumption of assigning blame is that this failure was a unique situation that could never have happened if the individual to blame had not been involved. Occasionally, there are cases where there is a clear bug in some software that leads to erroneous results. Fixing the bug in the code will “fix” the results, but even in that situation it’s not always clear that the bug is the ultimate cause of failure (although in this case it is the proximate cause). The case study presented in the next chapter is useful for thinking about what kinds of generalizable knowledge we can obtain from data analysis failures. This case is special because it had serious implications and large parts of it played out in public. While we likely will never know all of the details, we know enough to have a meaningful discussion about the lessons learned. "],["case-study-developing-a-genomic-predictor.html", "Chapter 4 Case Study: Developing a Genomic Predictor 4.1 Initial Lessons 4.2 New Information Appears 4.3 Lessons Learned 4.4 Summary", " Chapter 4 Case Study: Developing a Genomic Predictor The case study presented here took place largely at Duke University and has at various times been referred to as the “Duke Saga”. At a high level, it involved a research group that claimed to have developed a genomic signature that predicted whether a person would respond to cancer chemotherapy. When others tried to reproduce the predictive results of the genomic signature, they were unable to do so. A brief time line is given below: In 2006, Nature Medicine published a paper by Anil Potti and colleagues titled “Genomic signature to guide the use of chemotherapeutics”. The paper claimed to have developed a classifier based on applying microarray technology to cell lines maintained by the National Cancer Institute (NCI). They claimed the classifier could determine which patients would respond to chemotherapy treatment. Keith Baggerly and Kevin Coombes, two biostatisticians at the MD Anderson Cancer Center were inundated by requests from (justifiably) excited colleagues who wanted to use this technology. Baggerly and Coombes attempted to reproduce the results using the published description but were unable to do so. They were able to reproduce certain results in the paper after deliberately introducing a series of errors into the data analysis. Since this initial incident, a number of other papers from the same lab were scrutinized and numerous errors in analyses were found, many that one might consider basic data handling and wrangling mistakes. In addition, Baggerly and Coombes found circumstantial evidence of deliberate fraud, such as claiming that certain genes were critical to a classifier even though those genes are not included in the microarray claimed to have been used. Clinical trials were started at Duke where patients were randomized into different arms of the trial based on the flawed techniques developed by Potti. After numerous scientists wrote a letter to NCI director Harold Varmus, Duke suspended the trials to investigate the situation. An internal Duke panel eventually cleared Potti and colleagues of any wrongdoing and restarted the trials. Time passes and eventually it is discovered by the The Cancer Letter that Potti lied on an application for federal funding about once being a Rhodes Scholar. Eventually, the trials were stopped, but only after much public scrutiny and a series of lawsuits. There is quite a bit more detail to this story, which played out over many years. If you want to hear more about this you can hear about it from Keith Baggerly himself in this nice lecture. This saga has been a difficult one to understand from the perspective of drawing generalizable lessons. While it’s fascinating because of the sheer number of problems that occurred, it’s not necessarily clear what intervention could be taken to prevent a similar episode from happening in the future. Problems with reproducibility clearly played a role here given that Baggerly and Coombes were initially unable to reproduce any of the original analyses. 4.1 Initial Lessons While the details of the Duke Saga were at time astonishing, it is difficult to draw any conclusion about what actually went wrong and what approach should be taken to prevent something like this from happening again. Most people on the outside were just speculating about what could have happened and the people who really would know the details weren’t talking very much. Here’s what most people seemed to take away from the publicly available information about the saga: Reproducibility. There was definitely a reproducible research angle to this saga, in that the analyses that were conducted lacked transparency. There was only sketchy code that was published along with paper and data were not immediately available. However, in a strange sense, much of what came to light did so because the work was ultimately partially reproducible. That is in fact how Baggerly and Coombes discovered all the problems. They were able to reproduce the findings after deliberately introducing mistakes in the data. If one went back in time and magically forced everyone in the lab to use R Markdown or Juypter Notebooks, it’s not clear how that would have prevented anything. For starters, everyone within the team had access to the analyses and the data. It’s possible that people outside the team might have discovered problems sooner if the work had been completely reproducible, but Baggerly and Coombes figured things out relatively quickly. Also, that is besides the point: We should not depend on people outside the research team as a primary defense against data analytic failure. It’s not clear that reproducibility is one of the lessons learned from this saga because it’s not clear that it would have made a difference. Expertise. The basic narrative explaining this saga was that the data analyses were poorly done. Statisticians in particular have focused on the use of proprietary software, non-reproducible workflows (like pointing and clicking in Excel), and incorrect application of otherwise sound statistical methodology (e.g. cross validation). Perhaps if better-trained people had been doing the analyses, none of this would have happened. Perhaps genomic analyses are too complicated for the traditionally trained laboratory scientist. The idea is that this kind of work is “hard to do” and that you need better people (or improve existing people). That is the gist of the summary in this segment from the television show 60 Minutes on the entire saga. Individual behavior. Anil Potti, the principal investigator of the study, was eventually fired from Duke over this scandal and most would agree with that decision. If Duke had fired him 10 years ago, then perhaps yes, this research would not have happened at Duke, but it might have happened somewhere else, or it might have happened at Duke but with a different principal investigator. So while Dr. Potti was ultimately responsible for the analyses, his firing does not provide a useful “lesson learned”. 4.2 New Information Appears In January 2015, The Cancer Letter published a blockbuster memo written by Bradford Perez, who in 2008 was a medical student trainee in the Potti lab. He saw what was going on in the lab and recognized its shoddiness. Problems that Baggerly and Coombes had to essentially reverse engineer, Perez saw first hand and immediately recognized them as serious. In fact, in 2008 he wrote a memo to the leadership of his institute describing some of those problems: “Fifty-nine cell line samples with mRNA expression data…were split in half to designate sensitive and resistant phenotypes. Then in developing the model, only those samples which fit the model best in cross validation were included. Over half of the original samples were removed…. This was an incredibly biased approach which does little more than give the appearance of a successful cross validation.” [emphasis added] He further wrote, At this point, I believe that the situation is serious enough that all further analysis should be stopped to evaluate what is known about each predictor and it should be reconsidered which are appropriate to continue using and under what circumstances…. I would argue that at this point nothing…should be taken for granted. All claims of predictor validations should be independently and blindly performed.” The memo was ignored by Institute leadership. Nothing was stopped and nothing was changed at the time. Perez eventually took his name off a series of papers and left the lab. 4.3 Lessons Learned The Perez memo is critical because it fundamentally changes the narrative about what went wrong in this entire saga. Yes, genomic analyses are “hard to do” but clearly there was expertise in the lab to recognize that difficulty and to recognize when statistical methods were being incorrectly applied. The problem was not a lack of training, nor was it simply the result of a few honest data management mistakes here and there. The problem was a breakdown in communication and a total lack of trust between investigators and members of the data analytic team. Perez clearly felt uncomfortable raising these issues in the lab and wrote the memo knowing that he had “much to lose”. He thought the problem in the lab was that statistical methods were being misapplied, but the real problem in the lab was that he did not feel comfortable discussing it. A breakdown in the relationship between an analyst and an investigator is a serious data analytic problem. It’s possible to imagine an alternate scenario where a data analyst like Perez sees a problem with the way models are being developed or applied, mentions this to the principal investigator and has a detailed discussion, perhaps seeks outside expertise (e.g. from a statistician), and then modifies the procedure to fix the problem. This is a process that happens pretty much every day in many labs around the world. No data analysis is perfect from start to finish. Changes and course corrections are constantly made along the way. problems that can be traced to data collection can be raised with the principal investigator. When results are given to other investigators, sometimes the results don’t seem right to them and they will seek clarification. Mistakes can be fixed and results can be updated. When the relationships between an analyst and various members of the investigator team are strong and there is substantial trust between them, honest mistakes are just minor bumps in the road that can be uncovered, discussed, and fixed. When there is a breakdown in those relationships, the exact same mistakes are covered up, denied, and buried. A breakdown in the relationships between analysts and other investigators on the team generally cannot be fixed with a better statistical method, or a reproducible workflow, or open source software. Recognizing that this is the problem is difficult because often there is no easy solution. 4.4 Summary The data analytic lesson learned from the Duke Saga is that data analysts need to be allowed to say “stop”. But also, the ability to do so depends critically on the relationships between the analyst and members of the investigator team. If an analyst feels uncomfortable raising analytic issues with other members, then arguably all analyses done by the team are at risk. No amount of statistical expertise or tooling can fix this fundamental human problem. "],["about-the-authors.html", "About the Authors", " About the Authors These credits are based on our course contributors table guidelines.     Credits Names Pedagogy Lead Content Instructor(s) Roger Peng Technical Template Publishing Engineers Candace Savonen, Carrie Wright Publishing Maintenance Engineer Candace Savonen Technical Publishing Stylists Carrie Wright, Candace Savonen Package Developers (ottrpal) Candace Savonen, John Muschelli, Carrie Wright Art and Design Illustrator(s) Candace Savonen, Carrie Wright Funding Funder(s) NIH Grant R25GM141505   ## ─ Session info ─────────────────────────────────────────────────────────────── ## setting value ## version R version 4.0.2 (2020-06-22) ## os Ubuntu 20.04.5 LTS ## system x86_64, linux-gnu ## ui X11 ## language (EN) ## collate en_US.UTF-8 ## ctype en_US.UTF-8 ## tz Etc/UTC ## date 2023-04-11 ## ## ─ Packages ─────────────────────────────────────────────────────────────────── ## package * version date lib source ## assertthat 0.2.1 2019-03-21 [1] RSPM (R 4.0.5) ## bookdown * 0.24 2023-03-28 [1] Github (rstudio/bookdown@88bc4ea) ## bslib 0.4.2 2022-12-16 [1] CRAN (R 4.0.2) ## cachem 1.0.7 2023-02-24 [1] CRAN (R 4.0.2) ## callr 3.5.0 2020-10-08 [1] RSPM (R 4.0.2) ## cli 3.6.1 2023-03-23 [1] CRAN (R 4.0.2) ## crayon 1.3.4 2017-09-16 [1] RSPM (R 4.0.0) ## desc 1.2.0 2018-05-01 [1] RSPM (R 4.0.3) ## devtools 2.3.2 2020-09-18 [1] RSPM (R 4.0.3) ## digest 0.6.25 2020-02-23 [1] RSPM (R 4.0.0) ## ellipsis 0.3.1 2020-05-15 [1] RSPM (R 4.0.3) ## evaluate 0.20 2023-01-17 [1] CRAN (R 4.0.2) ## fastmap 1.1.1 2023-02-24 [1] CRAN (R 4.0.2) ## fs 1.5.0 2020-07-31 [1] RSPM (R 4.0.3) ## glue 1.4.2 2020-08-27 [1] RSPM (R 4.0.5) ## htmltools 0.5.5 2023-03-23 [1] CRAN (R 4.0.2) ## jquerylib 0.1.4 2021-04-26 [1] CRAN (R 4.0.2) ## jsonlite 1.7.1 2020-09-07 [1] RSPM (R 4.0.2) ## knitr 1.33 2023-03-28 [1] Github (yihui/knitr@a1052d1) ## magrittr 2.0.3 2022-03-30 [1] CRAN (R 4.0.2) ## memoise 2.0.1 2021-11-26 [1] CRAN (R 4.0.2) ## pkgbuild 1.1.0 2020-07-13 [1] RSPM (R 4.0.2) ## pkgload 1.1.0 2020-05-29 [1] RSPM (R 4.0.3) ## prettyunits 1.1.1 2020-01-24 [1] RSPM (R 4.0.3) ## processx 3.4.4 2020-09-03 [1] RSPM (R 4.0.2) ## ps 1.4.0 2020-10-07 [1] RSPM (R 4.0.2) ## R6 2.4.1 2019-11-12 [1] RSPM (R 4.0.0) ## remotes 2.2.0 2020-07-21 [1] RSPM (R 4.0.3) ## rlang 1.1.0 2023-03-14 [1] CRAN (R 4.0.2) ## rmarkdown 2.10 2023-03-28 [1] Github (rstudio/rmarkdown@02d3c25) ## rprojroot 2.0.3 2022-04-02 [1] CRAN (R 4.0.2) ## sass 0.4.5 2023-01-24 [1] CRAN (R 4.0.2) ## sessioninfo 1.1.1 2018-11-05 [1] RSPM (R 4.0.3) ## stringi 1.5.3 2020-09-09 [1] RSPM (R 4.0.3) ## stringr 1.4.0 2019-02-10 [1] RSPM (R 4.0.3) ## testthat 3.0.1 2023-03-28 [1] Github (R-lib/testthat@e99155a) ## usethis 1.6.3 2020-09-17 [1] RSPM (R 4.0.2) ## withr 2.3.0 2020-09-22 [1] RSPM (R 4.0.2) ## xfun 0.26 2023-03-28 [1] Github (yihui/xfun@74c2a66) ## yaml 2.2.1 2020-02-01 [1] RSPM (R 4.0.3) ## ## [1] /usr/local/lib/R/site-library ## [2] /usr/local/lib/R/library "],["references.html", "References", " References "],["404.html", "Page not found", " Page not found The page you requested cannot be found (perhaps it was moved or renamed). You may want to try searching to find the page's new location, or use the table of contents to find the page you are looking for. "]]