ezr.h2o_compare_regr: Title H2o Regression Model Compare

In jmp1989/easyr: Helpful wrappers for common EDA, Data Manipulation, & Modeling

Description

This function makes it easier to compare multiple models side by side. It takes an input vector of model names and returns the performance metrics for validation dataset, training dataset, test dataset, xval dataset as desired.

Usage

ezr.h2o_compare_regr(vector_of_models, compare_train = FALSE,
  compare_test = TRUE, test_dataset = NULL, compare_xval = FALSE,
  compare_valid = FALSE, return_full_model_params = FALSE)

Arguments

compare_train	Get the metrics for train? Defaults to False
compare_test	Get the metrics for test? Defaults to true. You need to provide a dataset to evaluate against!
test_dataset	The dataset that you will predict against.
compare_xval	Get the metrics for XVAL? Defaults to false.
compare_valid	Get the metrics for Validation dataset? Defautls to False
return_full_model_params	False by default. If FALSE returns only the tuning metrics that were used. If TRUE, will return everything.

Value

Returns a dataframe of the metrics chosen. if 'all' is chosen on the return_type, then a list of dataframes is returned.

jmp1989/easyr documentation built on May 20, 2019, 7:25 a.m.