The Bertin Classification Index — BCI • extracat

Computes the Bertin Classification Index for a contingency table of any dimensions.

BCI(x)

Arguments

x	A data matrix, table or array.

Details

The BCI is the Bertin Classification Criterion (BCC) normalized by the BCC value under independence.

Value

The criterion value.

See also

Examples

#for an unoptimized matrix we take the minimum of BCI(M) and BCI(M[,12:1])
M <-arsim(1000, c(12,12), 3)
min(BCI(M), BCI(M[,12:1]))
#> [1] 0.8308147

#an strongly related alternative (for two-way data)
kendalls(M)
#> [1] 0.1694221

M2 <- optile(M, iter = 100)
BCI(M2)
#> [1] 0.2248789
kendalls(M2)
#> [1] 0.7054944

M3 <-arsim(100000, c(12,13,15), 4,noise=0.2,shuffle=FALSE)
BCI(M3)
#> [1] 0.2919551