posterior predictive check for binomial glm with matrix response

[strengejacke]

Fixes #644
Fixes #646

[codecov]

Codecov Report

Merging #645 (e10d899) into main (5dc4f80) will increase coverage by 0.33%.
The diff coverage is 43.18%.

❗ Current head e10d899 differs from pull request most recent head 8b9b014. Consider uploading reports for the commit 8b9b014 to get more accurate results

@@            Coverage Diff             @@
##             main     #645      +/-   ##
==========================================
+ Coverage   55.08%   55.41%   +0.33%     
==========================================
  Files          84       84              
  Lines        5884     5894      +10     
==========================================
+ Hits         3241     3266      +25     
+ Misses       2643     2628      -15     

Files	Coverage Δ
R/binned_residuals.R	75.28% <100.00%> (+1.16%)	⬆️
R/looic.R	37.50% <ø> (ø)
R/check_model.R	31.29% <0.00%> (ø)
R/check_predictions.R	39.74% <33.33%> (+12.41%)	⬆️
R/r2_bayes.R	53.00% <40.00%> (ø)

[strengejacke]

@IndrajeetPatil lintr complains about:

Warning: file=R/check_predictions.R,line=434,col=30,[unnecessary_lambda_linter] Pass exp directly as a symbol to lapply() instead of wrapping it in an unnecessary anonymous function. For example, prefer lapply(DF, sum) to lapply(DF, function(x) sum(x)).

The related code line is:

sims[] <- lapply(sims, function(i) exp(i) - plus_minus)

Can I really re-write that line w/o using function(), or is this a false-positive?

[IndrajeetPatil]

@strengejacke That's a false positive.

Can you please report it here: https://github.com/r-lib/lintr/issues

Thanks!