Skip to content

Commit

Permalink
deploy: 917e4d2
Browse files Browse the repository at this point in the history
  • Loading branch information
@ljchang
ljchang committed Apr 3, 2024
1 parent 59e398d commit 41ab219
Show file tree
Hide file tree
Showing 13 changed files with 34 additions and 42 deletions.
6 changes: 2 additions & 4 deletions _sources/content/GLM_Single_Subject_Model.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1112,13 +1112,11 @@
]
},
{
"cell_type": "code",
"execution_count": null,
"cell_type": "markdown",
"metadata": {},
"outputs": [],
"source": [
"```{note}\n",
"Multicollineary as discussed above can lead to unstable estimates of the regression coefficients, particularly for the regressors of interest. Multicollinearity between covariates of no interest tends to be less of a problem because we generally are just interested in explaining noise in our model and are rarely interested in interpreting the individual covariate beta coefficients. However, in cases of extreme collinearity, the model may become rank deficient, which can lead to difficulty with even fitting the model. \n",
"As discussed above, multicollineary can lead to unstable estimates of the regression coefficients, which is particularly important to keep in mind for regressors of interest. Multicollinearity between covariates of no interest tends to be less of a problem because we generally are just interested in explaining noise in our model and are rarely interested in interpreting the individual covariate beta coefficients. However, in cases of extreme collinearity, the model may become rank deficient, which can lead to difficulty with even fitting the model. \n",
"\n",
"A simple fix to this problem is to use the `.clean()` method. This method will remove any columns that are perfectly collinear with other columns in the design matrix.\n",
"```"
Expand Down
2 changes: 1 addition & 1 deletion content/Download_Data.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -740,7 +740,7 @@ <h4>Windows Path Separators<a class="headerlink" href="#windows-path-separators"
</section>
<section id="download-data-with-datalad">
<h2>Download Data with DataLad<a class="headerlink" href="#download-data-with-datalad" title="Permalink to this heading">#</a></h2>
<p>The Pinel localizer dataset can be accessed at the following location https://gin.g-node.org/ljchang/Localizer/. To download the Localizer dataset run <code class="docutils literal notranslate"><span class="pre">datalad</span> <span class="pre">install</span> <span class="pre">https://gin.g-node.org/ljchang/Localizer</span></code> in a terminal in the location where you would like to install the dataset. Don’t forget to change the directory to a folder on your local computer. The full dataset is approximately 42gb.</p>
<p>The Pinel localizer dataset can be accessed at the following location <a class="reference external" href="https://gin.g-node.org/ljchang/Localizer/">https://gin.g-node.org/ljchang/Localizer/</a>. To download the Localizer dataset run <code class="docutils literal notranslate"><span class="pre">datalad</span> <span class="pre">install</span> <span class="pre">https://gin.g-node.org/ljchang/Localizer</span></code> in a terminal in the location where you would like to install the dataset. Don’t forget to change the directory to a folder on your local computer. The full dataset is approximately 42gb.</p>
<p>You can run this from the notebook using the <code class="docutils literal notranslate"><span class="pre">!</span></code> cell magic.</p>
<div class="cell docutils container">
<div class="cell_input docutils container">
Expand Down
14 changes: 4 additions & 10 deletions content/GLM_Single_Subject_Model.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1077,16 +1077,10 @@ <h3>Noise Covariates<a class="headerlink" href="#noise-covariates" title="Permal
<img alt="../_images/b398f2671618d411b8f16f9ba60298a299f41645c99b821341feaf6155ba6ddc.png" src="../_images/b398f2671618d411b8f16f9ba60298a299f41645c99b821341feaf6155ba6ddc.png" />
</div>
</div>
<div class="cell docutils container">
<div class="cell_input docutils container">
<div class="highlight-ipython3 notranslate"><div class="highlight"><pre><span></span>```{note}
Multicollineary as discussed above can lead to unstable estimates of the regression coefficients, particularly for the regressors of interest. Multicollinearity between covariates of no interest tends to be less of a problem because we generally are just interested in explaining noise in our model and are rarely interested in interpreting the individual covariate beta coefficients. However, in cases of extreme collinearity, the model may become rank deficient, which can lead to difficulty with even fitting the model.

A simple fix to this problem is to use the `.clean()` method. This method will remove any columns that are perfectly collinear with other columns in the design matrix.
```
</pre></div>
</div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>As discussed above, multicollineary can lead to unstable estimates of the regression coefficients, which is particularly important to keep in mind for regressors of interest. Multicollinearity between covariates of no interest tends to be less of a problem because we generally are just interested in explaining noise in our model and are rarely interested in interpreting the individual covariate beta coefficients. However, in cases of extreme collinearity, the model may become rank deficient, which can lead to difficulty with even fitting the model.</p>
<p>A simple fix to this problem is to use the <code class="docutils literal notranslate"><span class="pre">.clean()</span></code> method. This method will remove any columns that are perfectly collinear with other columns in the design matrix.</p>
</div>
</section>
</section>
Expand Down
2 changes: 1 addition & 1 deletion content/Instructors.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -455,7 +455,7 @@ <h2>Professors<a class="headerlink" href="#professors" title="Permalink to this
<a class="reference internal image-reference" href="../_images/finn.jpg"><img alt="../_images/finn.jpg" src="../_images/finn.jpg" style="width: 200px;" /></a>
<p><a class="reference external" href="https://esfinn.github.io/">Emily Finn</a>, PhD (Winter 2021) is an Assistant Professor of Psychological and Brain Sciences at Dartmouth College and directs the <a class="reference external" href="http://thefinnlab.github.io/">Functional Imaging &amp; Naturalistic Neuroscience (FINN) Lab</a>. She completed a BA in linguistics at Yale University and a PhD in neuroscience, also at Yale. She then did her postdoctoral training in the Section on Functional Imaging Methods in the Laboratory of Brain and Cognition and the National Institute of Mental Health. Her research is focused on individual variability in brain activity and behavior, especially as it relates to appraisal of ambiguous information under naturalistic conditions. Professor Finn is committed to the ideals of open science, including data and code sharing (see examples <a class="reference external" href="https://openneuro.org/datasets/ds001338">here</a>, <a class="reference external" href="https://github.com/esfinn/cpm_tutorial">here</a>, and <a class="reference external" href="https://github.com/esfinn/intersubj_rsa">here</a>, and to helping train other scientists in innovative new methods for neuroimaging data acquisition and analysis.</p>
<a class="reference internal image-reference" href="../_images/wager.jpg"><img alt="../_images/wager.jpg" src="../_images/wager.jpg" style="width: 200px;" /></a>
<p><a class="reference external" href="https://sites.dartmouth.edu/canlab/">Tor Wager</a>, PhD (Spring 2021) is the Diana L. Taylor Distinguished Professor in Neuroscience at Dartmouth College. He received his Ph.D. from the University of Michigan in Cognitive Psychology in 2003, and served as an Assistant (2004-2008) and Associate Professor (2009) at Columbia University, and as Associate (2010-2014) and Full Professor (2014-2019) at the University of Colorado, Boulder. Since 2004, he has directed the <a class="reference external" href="https://sites.dartmouth.edu/canlab/">Cognitive and Affective Neuroscience laboratory</a>, a research lab devoted to work on the neurophysiology of affective processes—pain, emotion, stress, and empathy—and how they are shaped by cognitive and social influences. Dr. Wager and his lab are also dedicated to developing analysis methods for functional neuroimaging and sharing ideas, tools, and scientific data with the scientific community and public. See https://canlab.github.io for papers, data, tools, and code.</p>
<p><a class="reference external" href="https://sites.dartmouth.edu/canlab/">Tor Wager</a>, PhD (Spring 2021) is the Diana L. Taylor Distinguished Professor in Neuroscience at Dartmouth College. He received his Ph.D. from the University of Michigan in Cognitive Psychology in 2003, and served as an Assistant (2004-2008) and Associate Professor (2009) at Columbia University, and as Associate (2010-2014) and Full Professor (2014-2019) at the University of Colorado, Boulder. Since 2004, he has directed the <a class="reference external" href="https://sites.dartmouth.edu/canlab/">Cognitive and Affective Neuroscience laboratory</a>, a research lab devoted to work on the neurophysiology of affective processes—pain, emotion, stress, and empathy—and how they are shaped by cognitive and social influences. Dr. Wager and his lab are also dedicated to developing analysis methods for functional neuroimaging and sharing ideas, tools, and scientific data with the scientific community and public. See <a class="reference external" href="https://canlab.github.io">https://canlab.github.io</a> for papers, data, tools, and code.</p>
<a class="reference internal image-reference" href="../_images/huckins.jpg"><img alt="../_images/huckins.jpg" src="../_images/huckins.jpg" style="width: 200px;" /></a>
<p><a class="reference external" href="https://mtnhuck.github.io/">Jeremy Huckins</a>, PhD (Fall 2019) is a Lecturer and Post-Doctoral researcher in the department of Psychological and Brain Sciences at Dartmouth College. He completed a BA in Neuroscience at Bowdoin College, worked with as a researcher with the <a class="reference external" href="https://king.med.harvard.edu/">King Lab</a> at Harvard Medical School then completed a PhD in Experimental and Molecular Medicine at Dartmouth College. His current research program is focused on gaining insights into mental health using fMRI and mobile smartphone sensing.</p>
</section>
Expand Down
24 changes: 12 additions & 12 deletions content/Introduction_to_Discovery.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -566,11 +566,11 @@ <h1>High Performance Computing at Dartmouth<a class="headerlink" href="#high-per
<p><em>Written by Courtney Jiminez &amp; Luke Chang</em></p>
<section id="what-is-high-performance-computing">
<h2>What is High Performance Computing?<a class="headerlink" href="#what-is-high-performance-computing" title="Permalink to this heading">#</a></h2>
<p>https://www.youtube.com/watch?v=nIBu1EFYmBU 00:13 - 01:21</p>
<p><a class="reference external" href="https://www.youtube.com/watch?v=nIBu1EFYmBU">https://www.youtube.com/watch?v=nIBu1EFYmBU</a> 00:13 - 01:21</p>
</section>
<section id="hpcs-at-dartmouth">
<h2>HPCs at Dartmouth<a class="headerlink" href="#hpcs-at-dartmouth" title="Permalink to this heading">#</a></h2>
<p>https://rc.dartmouth.edu/index.php/discoveryhpc/</p>
<p><a class="reference external" href="https://rc.dartmouth.edu/index.php/discoveryhpc/">https://rc.dartmouth.edu/index.php/discoveryhpc/</a></p>
<section id="discovery-cluster">
<h3>Discovery Cluster<a class="headerlink" href="#discovery-cluster" title="Permalink to this heading">#</a></h3>
<p>Where is Discovery located? In the basement of Baker Berry! In a huge room full of servers. Racks of computers. Each computer is a node - newer nodes have 32 cores with 8GB per core. Used by all Departments (PBS, CS, Physics &amp; Astronomy, etc.)</p>
Expand All @@ -584,16 +584,16 @@ <h3>System Layout<a class="headerlink" href="#system-layout" title="Permalink to
<section id="research-computing-team">
<h3>Research Computing Team<a class="headerlink" href="#research-computing-team" title="Permalink to this heading">#</a></h3>
<p>Centralized high-performance computing system group.</p>
<p>services.dartmouth.edu - where help/request forms are for research computing (try searching for Discovery, SLURM, etc.) Locally written documentation for things that people have problems with.</p>
<p><a class="reference external" href="http://services.dartmouth.edu">services.dartmouth.edu</a> - where help/request forms are for research computing (try searching for Discovery, SLURM, etc.) Locally written documentation for things that people have problems with.</p>
<p>Simple, basic examples of what you can do with the scheduler.</p>
<p>If you have a Discovery question, should go to services.dartmouth.edu, search for a key word, see if it’s already written up. If you have that question, likely a bunch of others did as well, and RC has tried to answer it within the Dartmouth context.</p>
<p>research.computing&#64;dartmouth.edu with questions.</p>
<p>https://services.dartmouth.edu</p>
<p>If you have a Discovery question, should go to <a class="reference external" href="http://services.dartmouth.edu">services.dartmouth.edu</a>, search for a key word, see if it’s already written up. If you have that question, likely a bunch of others did as well, and RC has tried to answer it within the Dartmouth context.</p>
<p><a class="reference external" href="mailto:research&#46;computing&#37;&#52;&#48;dartmouth&#46;edu">research<span>&#46;</span>computing<span>&#64;</span>dartmouth<span>&#46;</span>edu</a> with questions.</p>
<p><a class="reference external" href="https://services.dartmouth.edu">https://services.dartmouth.edu</a></p>
</section>
<section id="dbic-resources-on-discovery">
<h3>DBIC Resources on Discovery<a class="headerlink" href="#dbic-resources-on-discovery" title="Permalink to this heading">#</a></h3>
<p>DBIC owns 15 nodes (16 cores?) on Discovery. Acess to ~1200 CPUs (Luke, what is this x5 multiplier? How do we go from 15x16=240 to 1200 CPUs?)</p>
<p>DBIC Node (Ndoli): high-RAM node (1.5 TB), 20-24 CPUs. Can only access it by directly logging in. UserID&#64;ndoli.dartmouth.edu (not available on scheduler).</p>
<p>DBIC Node (Ndoli): high-RAM node (1.5 TB), 20-24 CPUs. Can only access it by directly logging in. <a class="reference external" href="mailto:UserID&#37;&#52;&#48;ndoli&#46;dartmouth&#46;edu">UserID<span>&#64;</span>ndoli<span>&#46;</span>dartmouth<span>&#46;</span>edu</a> (not available on scheduler).</p>
</section>
<section id="pbs-hpcs">
<h3>PBS HPCs<a class="headerlink" href="#pbs-hpcs" title="Permalink to this heading">#</a></h3>
Expand Down Expand Up @@ -639,7 +639,7 @@ <h3>ACL Permissions<a class="headerlink" href="#acl-permissions" title="Permalin
<p>see AskPBS</p>
<p>know what they are when you run into a problem and google it LOL</p>
<p>can also just ask research computing to update your permissions</p>
<p>https://services.dartmouth.edu/TDClient/1806/Portal/KB/ArticleDet?ID=88459</p>
<p><a class="reference external" href="https://services.dartmouth.edu/TDClient/1806/Portal/KB/ArticleDet?ID=88459">https://services.dartmouth.edu/TDClient/1806/Portal/KB/ArticleDet?ID=88459</a></p>
</section>
</section>
<section id="where-is-my-data-stored-on-discovery">
Expand Down Expand Up @@ -736,8 +736,8 @@ <h3>Running Batch Jobs<a class="headerlink" href="#running-batch-jobs" title="Pe
<p>TASKS/NODES (PROCESSES/NODE) - if your job is a call to one program (even if multi-threaded), it just does one task - only need 1 core here (which is the default). Never need more than 1 unless running an MPI job that knows how to run in parallel.</p>
<p>CPUs/TASK (CORES/PROCESS) - can request more than 1 for multi-threaded/multi-process jobs. (e.g. fMRIprep can run multi-threaded, can specify 16 cpus/threads)</p>
<p>RAM - hardest thing to estimate. Ideally you run a similar job in an unrestricted system and use top to see how much memory it allocates for the job, then request similar amount.</p>
<p>Walltime (doesn’t hurt too much to overestimate but don’t get wild LOL - will keep others’ jobs from starting). Noticed your job isn’t finished but walltime is running out? Can extend walltime (for up to 30 days!) but need to ask research computing to do this. Email research.computing&#64;dartmouth.edu. You can use the SBATCH qnotify directive to let you know a certain amont of time before your job terminates.</p>
<p>Other SBATCH directives: check out this link. https://slurm.schedmd.com/sbatch.html</p>
<p>Walltime (doesn’t hurt too much to overestimate but don’t get wild LOL - will keep others’ jobs from starting). Noticed your job isn’t finished but walltime is running out? Can extend walltime (for up to 30 days!) but need to ask research computing to do this. Email <a class="reference external" href="mailto:research&#46;computing&#37;&#52;&#48;dartmouth&#46;edu">research<span>&#46;</span>computing<span>&#64;</span>dartmouth<span>&#46;</span>edu</a>. You can use the SBATCH qnotify directive to let you know a certain amont of time before your job terminates.</p>
<p>Other SBATCH directives: check out this link. <a class="reference external" href="https://slurm.schedmd.com/sbatch.html">https://slurm.schedmd.com/sbatch.html</a></p>
</section>
<section id="arrays-running-multiple-batch-jobs-at-once">
<h3>Arrays: Running Multiple Batch Jobs at Once<a class="headerlink" href="#arrays-running-multiple-batch-jobs-at-once" title="Permalink to this heading">#</a></h3>
Expand All @@ -764,7 +764,7 @@ <h3>Scheduler Etiquette<a class="headerlink" href="#scheduler-etiquette" title="
</section>
<section id="can-i-use-jupyter-notebook-on-discovery">
<h2>Can I use Jupyter Notebook on Discovery?<a class="headerlink" href="#can-i-use-jupyter-notebook-on-discovery" title="Permalink to this heading">#</a></h2>
<p>AskPBS.org link</p>
<p><a class="reference external" href="http://AskPBS.org">AskPBS.org</a> link</p>
</section>
<section id="have-additional-discovery-questions">
<h2>Have Additional Discovery Questions?<a class="headerlink" href="#have-additional-discovery-questions" title="Permalink to this heading">#</a></h2>
Expand All @@ -773,7 +773,7 @@ <h3>AskPBS<a class="headerlink" href="#askpbs" title="Permalink to this heading"
</section>
<section id="dbic-handbook">
<h3>DBIC Handbook<a class="headerlink" href="#dbic-handbook" title="Permalink to this heading">#</a></h3>
<p>https://dbic-handbook.readthedocs.io/en/latest/discovery.html</p>
<p><a class="reference external" href="https://dbic-handbook.readthedocs.io/en/latest/discovery.html">https://dbic-handbook.readthedocs.io/en/latest/discovery.html</a></p>
</section>
</section>
<section id="references">
Expand Down
2 changes: 1 addition & 1 deletion content/Introduction_to_Pandas.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2243,7 +2243,7 @@ <h3>Create New Columns<a class="headerlink" href="#create-new-columns" title="Pe
</section>
<section id="indexing-and-slicing-data">
<h2>Indexing and slicing Data<a class="headerlink" href="#indexing-and-slicing-data" title="Permalink to this heading">#</a></h2>
<p>Indexing in Pandas can be tricky. There are many ways to index in pandas, for this tutorial we will focus on four: loc, iloc, boolean, and indexing numpy values. For a more in depth overview see Jake Vanderplas’s tutorial](https://github.com/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/03.02-Data-Indexing-and-Selection.ipynb), where he also covers more advanced topics, such as <a class="reference external" href="https://github.com/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/03.05-Hierarchical-Indexing.ipynb">hierarchical indexing</a>.</p>
<p>Indexing in Pandas can be tricky. There are many ways to index in pandas, for this tutorial we will focus on four: loc, iloc, boolean, and indexing numpy values. For a more in depth overview see Jake Vanderplas’s tutorial](<a class="github reference external" href="https://github.com/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/03.02-Data-Indexing-and-Selection.ipynb">jakevdp/PythonDataScienceHandbook</a>), where he also covers more advanced topics, such as <a class="reference external" href="https://github.com/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/03.05-Hierarchical-Indexing.ipynb">hierarchical indexing</a>.</p>
<section id="indexing-with-keys">
<h3>Indexing with Keys<a class="headerlink" href="#indexing-with-keys" title="Permalink to this heading">#</a></h3>
<p>First, we will cover indexing with keys using the <code class="docutils literal notranslate"><span class="pre">.loc</span></code> method. This method references the explicit index with a key name. It works for both index names and also column names. Note that often the keys for rows are integers by default.</p>
Expand Down
Loading

0 comments on commit 41ab219

Please sign in to comment.