calcNested: Calculate 'plNested' Performance

In tpq/exprso: Rapid Deployment of Machine Learning Algorithms

Description

calcNested calculates a single performance measure for the results of a plNested function call.

Usage

calcNested(pl, colBy = "valid.acc")

Arguments

pl	Specifies the ExprsPipeline object returned by plNested.
colBy	A character vector or string. Specifies column(s) to use when summarizing model performance. Listing multiple columns will calculate performance as a product of those listed performances.

Details

For each dataset split (i.e., bootstrap), calcNested averages the validation set performance for the "best" model (where "best" is defined as the model with the maximum "internal" cross-validation accuracy, max($train.plCV)). The validation set performance ultimately averaged depends on the supplied colBy argument.

Value

A numeric scalar. The cross-validation accuracy.

tpq/exprso documentation built on July 27, 2019, 8:44 a.m.