testsplit: Split data set into training and validation sets, stratifying...

In bwrc/astrid-r: Automatic STructure IDentification

Description

Split data set into training and validation sets, stratifying by groups. Of each class of n items, floor(fraction*n) go to training set and ceiling((1-fraction)*n) go to validation set.

Usage

testsplit(data, fraction = 0.5, size = NULL, replace = FALSE,
  stratify = "Class")

Arguments

data	The data matrix
fraction	The fraction of data to place in the training and testing set.
size	The amount of data to place in the training and testing set
replace	Should the splitting be with replacement. Default is FALSE.
stratify	Variable according to which stratification should be performed.

Value

Indices of the training set.

bwrc/astrid-r documentation built on May 13, 2019, 9:08 a.m.