Help: Error in yhats[[which.min(yhats_diff)]]

[Margarida-F]

Hi !

Can I have help to fix the error below?

Error in yhats[[which.min(yhats_diff)]] :
attempt to select less than one element in get1index

I only code this:

contribution <- broken(model = model, new_observation = df[1,features], data = train[,features])

and my model is a randomForest.

It seems that is trying to access to the zero position, and it doesn't exist.

Thanks, Margarida

[pbiecek]

Do you have a reproducible example that produces this error?
And please try the iBreakDown package (https://github.com/ModelOriented/iBreakDown) with break_down) function, it's a newer version so maybe this problem is already solved there

[Margarida-F]

First of all, thanks for you answer.

I suppose I can't show the data but I guarantee the variables are all right.
However, maybe the problem is the model...

I try this:

contribution <- break_down(model, new_observation = df[1,features], data = train[,features])
Error in colMeans(yhatpred) : 'x' must be numeric

And then, tried construct explainer first:

explainer <- explain(model,data=train[,features],y=train$Y)
Error in UseMethod("explain") :
no applicable method for 'explain' applied to an object of class "c('randomForest.formula', 'randomForest')"

My model is like:

Can you help me, please?

Thank you, very much ,
Margarida

[pbiecek]

Can you share devtools::session_info()?
It is hard to suggest anything without knowing on which versions of DALEX/breakDown you are working.
Try latest versions DALEX & iBreakDown from CRAN.

If this is not enough then try to change thy y variable from factor to numeric
(as Y == "1")

[Margarida-F]

Hi, again. Thank you very much you're trying to help me.

My Session Information is:

• Session info -----------------------------------------------------------------------------------------------------------------
  setting value
  version R version 3.5.2 (2018-12-20)
  os Windows >= 8 x64
  system x86_64, mingw32
  ui RStudio
  language (EN)
  collate Portuguese_Portugal.1252
  ctype Portuguese_Portugal.1252
  tz Europe/London
  date 2019-12-09

• Packages ---------------------------------------------------------------------------------------------------------------------
  package * version date lib source
  assertthat 0.2.1 2019-03-21 [1] CRAN (R 3.5.3)
  backports 1.1.5 2019-10-02 [1] CRAN (R 3.5.3)
  callr 3.3.2 2019-09-22 [1] CRAN (R 3.5.3)
  caret * 6.0-84 2019-04-27 [1] CRAN (R 3.5.3)
  class 7.3-14 2015-08-30 [2] CRAN (R 3.5.2)
  cli 1.1.0 2019-03-19 [1] CRAN (R 3.5.3)
  codetools 0.2-15 2016-10-05 [2] CRAN (R 3.5.2)
  colorspace 1.4-1 2019-03-18 [1] CRAN (R 3.5.3)
  crayon 1.3.4 2017-09-16 [1] CRAN (R 3.5.3)
  DALEX 0.4.9 2019-10-24 [1] CRAN (R 3.5.3)
  data.table * 1.12.2 2019-04-07 [1] CRAN (R 3.5.3)
  DBI 1.0.0 2018-05-02 [1] CRAN (R 3.5.3)
  dbplyr * 1.4.2 2019-06-17 [1] CRAN (R 3.5.3)
  desc 1.2.0 2018-05-01 [1] CRAN (R 3.5.3)
  devtools 2.2.1 2019-09-24 [1] CRAN (R 3.5.2)
  digest 0.6.23 2019-11-23 [1] CRAN (R 3.5.3)
  dplyr * 0.8.3 2019-07-04 [1] CRAN (R 3.5.3)
  ellipsis 0.3.0 2019-09-20 [1] CRAN (R 3.5.3)
  foreach 1.4.7 2019-07-27 [1] CRAN (R 3.5.3)
  fs 1.3.1 2019-05-06 [1] CRAN (R 3.5.3)
  generics 0.0.2 2018-11-29 [1] CRAN (R 3.5.3)
  ggplot2 * 3.2.1 2019-08-10 [1] CRAN (R 3.5.3)
  glue 1.3.1 2019-03-12 [1] CRAN (R 3.5.3)
  gower 0.2.1 2019-05-14 [1] CRAN (R 3.5.3)
  gtable 0.3.0 2019-03-25 [1] CRAN (R 3.5.3)
  iBreakDown 0.9.9 2019-08-26 [1] CRAN (R 3.5.3)
  ipred 0.9-9 2019-04-28 [1] CRAN (R 3.5.3)
  iterators 1.0.12 2019-07-26 [1] CRAN (R 3.5.3)
  lattice * 0.20-38 2018-11-04 [2] CRAN (R 3.5.2)
  lava 1.6.6 2019-08-01 [1] CRAN (R 3.5.3)
  lazyeval 0.2.2 2019-03-15 [1] CRAN (R 3.5.3)
  lubridate * 1.7.4 2018-04-11 [1] CRAN (R 3.5.3)
  magrittr 1.5 2014-11-22 [1] CRAN (R 3.5.3)
  MASS 7.3-51.1 2018-11-01 [2] CRAN (R 3.5.2)
  Matrix 1.2-15 2018-11-01 [2] CRAN (R 3.5.2)
  memoise 1.1.0 2017-04-21 [1] CRAN (R 3.5.3)
  ModelMetrics 1.2.2 2018-11-03 [1] CRAN (R 3.5.3)
  munsell 0.5.0 2018-06-12 [1] CRAN (R 3.5.3)
  nlme 3.1-137 2018-04-07 [2] CRAN (R 3.5.2)
  nnet 7.3-12 2016-02-02 [2] CRAN (R 3.5.2)
  pillar 1.4.2 2019-06-29 [1] CRAN (R 3.5.3)
  pkgbuild 1.0.5 2019-08-26 [1] CRAN (R 3.5.3)
  pkgconfig 2.0.3 2019-09-22 [1] CRAN (R 3.5.3)
  pkgload 1.0.2 2018-10-29 [1] CRAN (R 3.5.3)
  plyr 1.8.4 2016-06-08 [1] CRAN (R 3.5.3)
  prettyunits 1.0.2 2015-07-13 [1] CRAN (R 3.5.3)
  processx 3.4.1 2019-07-18 [1] CRAN (R 3.5.3)
  prodlim 2018.04.18 2018-04-18 [1] CRAN (R 3.5.3)
  ps 1.3.0 2018-12-21 [1] CRAN (R 3.5.3)
  purrr 0.3.2 2019-03-15 [1] CRAN (R 3.5.3)
  R6 2.4.0 2019-02-14 [1] CRAN (R 3.5.3)
  randomForest * 4.6-14 2018-03-25 [1] CRAN (R 3.5.3)
  Rcpp 1.0.3 2019-11-08 [1] CRAN (R 3.5.3)
  recipes 0.1.7 2019-09-15 [1] CRAN (R 3.5.2)
  remotes 2.1.0 2019-06-24 [1] CRAN (R 3.5.3)
  reshape2 1.4.3 2017-12-11 [1] CRAN (R 3.5.3)
  rlang 0.4.1 2019-10-24 [1] CRAN (R 3.5.3)
  rpart 4.1-13 2018-02-23 [2] CRAN (R 3.5.2)
  rprojroot 1.3-2 2018-01-03 [1] CRAN (R 3.5.3)
  rstudioapi 0.10 2019-03-19 [1] CRAN (R 3.5.2)
  scales 1.0.0 2018-08-09 [1] CRAN (R 3.5.3)
  sessioninfo 1.1.1 2018-11-05 [1] CRAN (R 3.5.3)
  stringi 1.4.3 2019-03-12 [1] CRAN (R 3.5.3)
  stringr 1.4.0 2019-02-10 [1] CRAN (R 3.5.3)
  survival 2.43-3 2018-11-26 [2] CRAN (R 3.5.2)
  testthat 2.2.1 2019-07-25 [1] CRAN (R 3.5.3)
  tibble 2.1.3 2019-06-06 [1] CRAN (R 3.5.3)
  tidyselect 0.2.5 2018-10-11 [1] CRAN (R 3.5.3)
  timeDate 3043.102 2018-02-21 [1] CRAN (R 3.5.3)
  usethis 1.5.1 2019-07-04 [1] CRAN (R 3.5.3)
  withr 2.1.2 2018-03-15 [1] CRAN (R 3.5.3)
  yaml 2.2.0 2018-07-25 [1] CRAN (R 3.5.3)

[1] C:/Users/Margarida Ferreira/Documents/R/win-library/3.5
[2] C:/Program Files/R/R-3.5.2/library

Change Y to numeric doesn't works because Y isn't an argument of break_down().

Should I download the code and run each code block to debug?

Thanks for everything,
Margarida

[Margarida-F]

Hi, again.

Just to close this issue:

I'm now using some code of iBreakDown. It's normal that it take too long (around 8 minutes) to calculate contributions of 13 variables and 1 row?

Thanks very much. I promise I close this issue before your answer.

Margarida

[pbiecek]

Hi,
busy time ar university, sorry for late response.

iBreakDown is much faster than breakDown.
For FIFA data (200k cases, ~100 variables) calculation of break down is below a minute, do for 13 variables it shall be seconds at most.

Here you will find how to use the most recent version of iBreakDown
https://pbiecek.github.io/PM_VEE/breakDown.html#BDR