chug: Chug a brew
In bcjaeger/ipa: Imputation for Predictive Analytics

Chug a brew

1	chug(brew, .fun = NULL, .fun_args = NULL)

brew

an ipa_brew object.

.fun

a function with at least three inputs: .trn .tst, and outcome. scrimp_mdl() will call your function as follows: .fun(.trn = train_imputed, .tst = test_imputed, outcome = outcome, ...), where ... is filled in by .fun_args. Generally, .fun should

develop a prediction model using .trn
create predicted values for observations in .tst
evaluates the predictions using a summary measure (e.g., R-squared, area underneath the ROC curve, Brier score, etc.).

See example below where a function using random forests is applied.

.fun_args

a named list with additional arguments for .fun.

an ipa_brew object with a column added to the wort containing the output from scrimp_mdl.

bcjaeger/ipa documentation built on May 7, 2020, 9:45 a.m.