Split your dataset into training, validation and test datasets.

This was largely inspired by train-test-split.

Install

npm install tvt-split
#or with Yarn
yarn add tvt-split

Documentation

Table of Contents

• trainValidationTestSplit
  • Parameters

trainValidationTestSplit

Split data into a training, validation and test set.

Parameters

• data Array Data
• train number Size of the training set (in percentage if train < 1)
• validation number Size of the validation set (in percentage if train < 1)
• seed number Seed of the Pseudo-Random Number Generation. (optional, default -1.1)
• indices boolean Indicates whether to return the indices of the provided array or the actual values (optional, default false)

Returns [Array, Array, Array] Split data

Example usage

const tvts = require('tvt-split');

const arr = ['one', 'two', 'three', 'four', 'five', 'six', 'seven', 'eight', 'nine', 'ten'];
const [train, validation, test] = tvts(arr, 0.8, .1, 1234);
console.log(train, validation, test);
// [ 'three', 'one', 'seven', 'eight', 'five', 'four', 'nine', 'six' ] [ 'two' ] [ 'ten' ]

const [trainIndices, validationIndices, testIndices] = tvts(arr, 0.8, .1, 1234, true);
console.log(trainIndices, validationIndices, testIndices);
// [ 2, 0, 6, 7, 4, 3, 8, 5, ] [ 1 ] [ 9 ]