ezr.h2o_get_thresholds: Get Threshold Metrics from h2o classification models

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

Description

Returns the entire table of threshold metrics for each model. Use the ezr.h2o_get_gridmodels to get the modelids out of grids. Determine whether or not these metrics should be at the training, xval, valid, or test data(newdata) level.

Usage

ezr.h2o_get_thresholds(h2omodels, xval = TRUE, newdata = NULL,
  valid = FALSE, train = FALSE)

Arguments

h2omodels	a vector of models such as c('id1','id2','id3')
xval	Use cross validation metrics?. This is true by default.
newdata	Use test data? Provide the test dataframe here
valid	Use validation data?
train	Use the training data?

Details

Only one metric should be true at a time

Value

Returns a dataframe of the concatenated metrics along with the model-id and algo.

Examples

  ezr.h2o_get_thresholds(h2omodels = your_vector_of_models, xval=FALSE, newdata='your_test_h2oframe' )

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