Given two partitions P and Q, cer(P, Q) measures how well
they agree,
the lower the better. It is rigorously defined as the proportion of pairwise
disagreements in the two partitions (i.e., how many, out of all the possible
couples of elements in the sample, are localized in the same cluster in one
partition and in a different one in the other partition).
Usage
cer(P, Q)
Value
The CER index, which is a number between 0 and 1, and also equal to
1 - Rand index (Rand, 1971), a popular measure of the goodness of a
clustering.
Arguments
P
first vector of cluster assignments (length n)
Q
second vector of cluster assignments (length n)
References
Rand, W. M. (1971). Objective criteria for the evaluation of
clustering methods. Journal of the American Statistical association, 66(336),
846-850.
set.seed(8988327)
x <- seq(0, 1, len = 500)
out <- generate.data.FV17(50, x)
result <- SparseFunClust(out$data, x, K = 2, do.alignment = FALSE)
cer(out$true.partition, result$labels)