splitDnnet: A function to split the 'dnnetInput' object into a list of...

In SkadiEye/deepTL: Deep Treatment Learning

A function to split the dnnetInput object into a list of two dnnetInput objects: one names train and the other named valid.

Description

A function to split the dnnetInput object into a list of two dnnetInput objects: one names train and the other named valid.

Usage

splitDnnet(object, split)

Arguments

object

A dnnetInput object.

split

A character, numeric variable or a numeric vector declaring a way to split the dnnetInput. If it's number between 0 and 1, all samples will be split into two subsets randomly, with the train containing such proportion of all samples and valid containing the rest. If split is a character and is "bootstrap", the train will be a bootstrap sample of the original data set and the valid will contain out-of-bag samples. If split is a vector of integers, the train will contain samples whose indice are in the vector, and valid will contain the rest.

Value

Returns a list of two dnnetInput objects.

See Also

dnnetInput-class

SkadiEye/deepTL documentation built on Nov. 17, 2022, 1:41 p.m.