compare_evaluate {drape}

Evaluate estimators by training nuisance functions on training set and evaluating them on test set.

Description

Evaluate estimators by training nuisance functions on training set and evaluating them on test set.

Usage

compare_evaluate(train, test, ex_setting, f_setting, regression, sm_bw_out)

Arguments

train

list containing vector of responses y and matrix of predictors X = (x,z).

test

list containing vector of responses y and matrix of predictors X = (x,z).

ex_setting

string "normal", "mixture2", "mixture3", "logistic", "t4", "401k".

f_setting

string "plm", "additive", "interaction".

regression

Optional fitted regression.

sm_bw_out

Output of cv_resmooth.

Value

list containing f, df, and score estimates evaluated on the test set.