It should be L squared (so R squared in the code)

[vguillemot]

sparseCorrespondenceAnalysis/R/pPMD.R

Line 185 in 7584941

r4 <- (n0inU + n0inV) / ((I + J) * (1:R))

[vguillemot]

It should be r4 <- (n0inU + n0inV) / ((I + J) * (1:R)^2)