EMMIXmfa (version 2.0.11)

ari: Computes adjusted Rand Index

Description

Computes adjusted Rand index.

Usage

ari(cls, hat_cls)

Value

Scaler specifying how closely two partitions agree.

Arguments

cls

A numeric or character vector of labels.

hat_cls

A numeric or character vector of labels same length as cls.

Details

Measures the agreement between two sets of partitions. The upper bound of 1 implies perfect agreement. The expected value is zero if the partitions are random.

References

Hubert L, and Arabie P (1985). Comparing Partitions. Journal of the Classification 2, 193--218.

See Also

minmis

Examples

set.seed(1984)
Y <- scale(iris[, -5])
model <- mfa(Y, g = 3, q = 3, nkmeans = 1, nrandom = 0)
#
ari(model$clust, iris[, 5])
#
minmis(model$clust, iris[, 5])