learn_curves: Learning curves

In drsimonj/pipelearner: Tidy machine learning pipelines

Description

Add learning curves (instructions to train on different proportions of each training set) to a pipelearner object.

Usage

learn_curves(pl, ...)

Arguments

pl	pipelearner object. See pipelearner
...	Non-zero proportions of training data to learn on. Will default to 1 if no values are provided.

Details

Will expect a pipelearner object, but will also accept a data frame for the parameter pl. In the case that a data frame is used, it will first coerce it to a pipelearner object via pipelearner.

These proportions are applied to cross-validation training sets using p_from_resample, meaning they will always slice the proportion starting the beginning. For example, a proportion of .5 will take the first 50% of rows from the training data (rounded to the nearest whole number). There is no point taking random subsamples of size p from the training data because it would require us to save these subsamples (taking up more space) despite the fact that the training data is already a random sample (see learn_cvpairs).

drsimonj/pipelearner documentation built on May 15, 2019, 2:53 p.m.