[R-package] added argument eval_train_metric to lgb.cv() (fixes #4911)

[mayer79]

Attempt to solve #4911

If we set eval_train_metric = TRUE in lgb.cv(), then we get output like

[1] "[50]: train's binary_logloss:0.248276+0.00109575 valid's binary_logloss:0.248655+0.00448549"

The resulting object "cvm" contains the information in the record_evals slot and the average training performance of the best round could be found by something like cvm$record_evals$train$binary_logloss$eval[[cvm$best_iter]].

There is no unit test yet, but that does not mean we shouldn't write one.

[jameslamb]

Thanks for picking this up!

Please do add a unit test testing this behavior. Let me know if you need help with it.

Please re-generate the documentation.

sh build-cran-package.sh --no-build-vignettes
R CMD INSTALL --with-keep.source ./lightgbm_3.3.1.99.tar.gz
cd R-package
Rscript -e "roxygen2::roxygenize(load = 'installed')"

...to fix issues like the following in CI (e.g. in this build)

Codoc mismatches from documentation object 'lgb.cv':
lgb.cv
  Code: function(params = list(), data, nrounds = 100L, nfold = 3L,
                 label = NULL, weight = NULL, obj = NULL, eval = NULL,
                 verbose = 1L, record = TRUE, eval_freq = 1L, showsd =
                 TRUE, eval_train_metric = FALSE, stratified = TRUE,
                 folds = NULL, init_model = NULL, colnames = NULL,
                 categorical_feature = NULL, early_stopping_rounds =
                 NULL, callbacks = list(), reset_data = FALSE,
                 serializable = TRUE)
  Docs: function(params = list(), data, nrounds = 100L, nfold = 3L,
                 label = NULL, weight = NULL, obj = NULL, eval = NULL,
                 verbose = 1L, record = TRUE, eval_freq = 1L, showsd =
                 TRUE, stratified = TRUE, folds = NULL, init_model =
                 NULL, colnames = NULL, categorical_feature = NULL,
                 early_stopping_rounds = NULL, callbacks = list(),
                 reset_data = FALSE, serializable = TRUE)
  Argument names in code not in docs:
    eval_train_metric
  Mismatches in argument names (first 3):
    Position: 13 Code: eval_train_metric Docs: stratified
    Position: 14 Code: stratified Docs: folds
    Position: 15 Code: folds Docs: init_model

[jameslamb]

thanks very much! The new test looks great! Just one more small suggestion.

[jameslamb]

looks great, thanks!

[jameslamb]

Thanks as always for the help @mayer79 !

[mayer79]

Thanks as always for the help @mayer79 !

Thanks go to you, sir :-)

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.