Class: TrainingSet

TrainingSet

Manage and format a set of recorded examples, maintain a rapid-mix compliant training set.

Constructor

new TrainingSet(inputDimensionopt, outputDimensionopt)

Parameters:
Name Type Attributes Default Description
inputDimension Number <optional>
 null

Input dimension (if null, is guessed from the first recorded element)
outputDimension Number <optional>
 null

Output dimension. (if null, is guessed from the first recorded element).
Source:
Example
import * as mano from 'mano-js/client';

const example = new mano.Example();
const trainingSet = new mano.TrainingSet();
const xmmProcessor = new mano.XmmProcesssor();

example.setLabel('test');
example.addElement([0, 1, 2, 3]);
const rapidMixJsonExample = example.toJSON();

trainingSet.addExample(rapidMixJsonExample);
const rapidMixJsonTrainingSet = trainingSet.toJSON();

xmmProcessor
  .train(rapidMixJsonTrainingSet)
  .then(() => { ... });

Members

length

Get the number of examples.

Source:

Methods

addExample(example)

Add an example to the training set.

Parameters:
Name Type Description
example JSON

A rapid-mix formatted example.
Source:

addTrainingSet(trainingSet)

Add all examples from another rapid-mix JSON training set.

Parameters:
Name Type Description
trainingSet JSON

A rapid-mix compliant training set.
Source:

clear()

Clear the training set.

Source:

getLabels() → {Array.String}

Return an array of the current training set labels.

Source:
Returns:
  • Sorted labels of the training set.
Type
Array.String

removeExampleAtIndex(index)

Remove example at index.

Parameters:
Name Type Description
index Number

The index of the example to remove.
Source:

removeExamplesByLabel(label)

Remove all examples of a certain label.

Parameters:
Name Type Description
label String

The label of the recordings to be removed.
Source:

setTrainingSet(trainingSet)

Initialize from another rapid-mix JSON training set. If null, clear the trainingSet.

Parameters:
Name Type Default Description
trainingSet JSON null

A rapid-mix compliant training set.
Source:

toJSON() → {JSON}

Return the rapid-mix compliant training set in JSON format.

Source:
Returns:
  • Training set.
Type
JSON