confusion: Confusion calculation

In wikiselev/clustools: Tools for cluster analysis of single cell high throughput data

Description

First test labels are matched to gold standard labels by using minWeightBipartiteMatching function from this package. Then when labels of labs and labs.known are of the same order, confusions between them are calculated. Confusion metrics introduced by myself:

Usage

confusion(labs, labs.known)

Arguments

labs	Cluster labels from test clustering
labs.known	Cluster labels from gold standard clustering

Details

1 - number-of-matching-labels/total-number-of-labels

where number-of-matching-labels - number of labels in each cluster group of labs that match to the labels of labs.known for the same cluster group; and total-number-of-labels - total number of labels of labs.known from the same cluster group.

Value

Vector containing confusions between labs and labs.known for each cluster group

Examples

res <- confusion(labs, labs.known)

wikiselev/clustools documentation built on May 4, 2019, 5:25 a.m.