find_best_models: Select top models from the find_model function

In jeroenvdhoven/datapiper: datapiper

Description

Select top models from the find_model function

Usage

find_best_models(train, find_model_result, metric, higher_is_better,
  per_model = F, top_n = 1, aggregate_func = NA)

Arguments

train	The training dataset
find_model_result	Result from the find_model function
metric	Target metric, as a string
higher_is_better	Logical indicating if the results should be sorted from high to low.
per_model	Logical indicating if we should take N models per model type or in total
top_n	The top n models to return
aggregate_func	Aggregation function to apply. Useful if you choose more than 1 model. Set this to NA to skip it.

Value

A pipeline to create predictions for new data

jeroenvdhoven/datapiper documentation built on July 14, 2019, 9:34 p.m.