Metrics homogeneity score

Development build. This page describes main, not a released package. The latest published Lodestar.Metrics is 0.3.0 — read its documentation.

Homogeneity.Score

Whether each cluster holds samples of a single class.

public static double Score(ReadOnlySpan<int> labelsTrue, ReadOnlySpan<int> labelsPred)

Parameters — labelsTrue is the reference partition and labelsPred the one being scored, one label per sample and the same length. The label values carry no meaning: only which samples share one does.

Returns — double in [0, 1]. 1 when every cluster is pure, 0 when the clustering says nothing about the classes.

Exceptions — ArgumentException when the two labellings disagree in length. An empty input is not an error: it scores 1.

Example — splitting a class in two keeps every cluster pure, and costs nothing here.

using Lodestar.Metrics;

int[] truth = [0, 0, 0, 1, 1, 1];
int[] split = [0, 0, 1, 2, 2, 2];

double pure = Homogeneity.Score(truth, split);   // => 1

Remarks — the example scores 1 while Completeness.Score on the same input scores less, and that asymmetry is the point of having both. This number answers "is any cluster mixed?" and nothing else; a clustering that splits every sample into its own cluster is perfectly homogeneous.

Applies to — net10.0, netstandard2.0.

See also — Completeness.Score, VMeasure.Score, the Python equivalence table.