grid_predict: For a grid of params, return list of model predictions.

In artichaud1/cook: A package for Tidy Hyperparameter Tuning

Description

This is a vectorized form of train_predict, vectorized over the param_grid.

Usage

grid_predict(split, recipe, param_grid, target_var, scoring_func,
  score_var = NULL, ..., verbosity, accept_func = NULL)

Arguments

split	A split A resampling split. See 'rsample' package.
recipe	Recipe. *Must have been prepped using 'retain = TRUE'. See 'recipe' package.
param_grid	List of list of parameters passed to the 'scoring_func' function.
target_var	Name of the target variable in the resample.
scoring_func	Your custom train/predict/score function.
score_var	Name of the score to use, in case 'scoring_func' returns a list of scores. If null (the default), will use the 1st score.
...	Optional params passed to scoring_func.
verbosity	If > 0, print paramset index.

Value

A list of data.frames with columns 'target' and 'predicted'.

artichaud1/cook documentation built on May 21, 2019, 9:23 a.m.