MetricStep: Metric Step

MetricStepR Documentation

Metric Step

Description

Metric Step

Metric Step

Super classes

rgudhi::PythonClass -> rgudhi::SKLearnClass -> MetricStep

Methods

Public methods

Inherited methods

Method apply()

Applies the class on a single persistence diagram and outputs the result.

Usage
MetricStep$apply(diag1, diag2)
Arguments
diag1

A 2-column tibble::tibble specifying a persistence diagram.

diag2

A 2-column tibble::tibble specifying a persistence diagram.

Returns

A numeric value storing the distance between the two input diagrams.


Method fit()

Fits the class on a sample of persistence diagrams.

Usage
MetricStep$fit(X, y = NULL)
Arguments
X

A list of 2-column tibble::tibbles specifying a sample of persistence diagrams.

y

An integer vector specifying persistence diagram labels (unused for now).

Returns

The class itself invisibly.


Method transform()

Applies the class on a sample of persistence diagrams.

Usage
MetricStep$transform(X)
Arguments
X

A list of 2-column tibble::tibbles specifying a sample of persistence diagrams.

Returns

A numeric matrix of shape n_\mathrm{out} \times n_\mathrm{in} storing the distances between the n_\mathrm{out} persistence diagrams passed to the ⁠$transform()⁠ method and the n_\mathrm{in} persistence diagrams passed to the ⁠$fit()⁠ method.


Method fit_transform()

Applies sequentially the ⁠$fit()⁠ and ⁠$transform()⁠ methods on a sample of persistence diagrams in a more efficient way than calling them directly.

Usage
MetricStep$fit_transform(X, y = NULL)
Arguments
X

A list of 2-column tibble::tibbles specifying a sample of persistence diagrams.

y

An integer vector specifying persistence diagram labels (unused for now).

Returns

A numeric matrix of shape n \times n storing the distance between the n persistence diagrams passed to both the ⁠$fit()⁠ and ⁠$transform()⁠ methods.


Method clone()

The objects of this class are cloneable with this method.

Usage
MetricStep$clone(deep = FALSE)
Arguments
deep

Whether to make a deep clone.

Author(s)

Mathieu Carrière


rgudhi documentation built on March 31, 2023, 11:38 p.m.