HBITM: Compute the HBITM similarity between two clusterings.

In clajusch/ClEvaR: Cluster Evaluation in R

Description

Compute the HBITM similarity between two clusterings.

Usage

HBITM(subject, query, weighted = TRUE)

Arguments

subject	Vector of reference cluster assignments.
query	Vector of cluster assignments for comparison.
weighted	Use weights? Defaults to TRUE.

Value

The (weighted) HBITM ("How blue is the matrix?") measure, a numeric vector of length 1.

Examples

a = c(rep("A", 1000), rep("B", 100), rep("C", 10))
b = c(rep("A", 500), rep("B", 595), rep("C", 15))
HBITM(subject = a, query = b, weighted = TRUE)
HBITM(subject = a, query = b, weighted = FALSE)

data(zeisel2018)
HBITM(subject = zeisel2018$cell_metadata$class,
      query = zeisel2018$cell_metadata$cluster_name,
      weighted = TRUE)
HBITM(subject = zeisel2018$cell_metadata$class,
      query = zeisel2018$cell_metadata$cluster_name,
      weighted = FALSE)

clajusch/ClEvaR documentation built on May 9, 2019, 4:55 a.m.