Calculate diversity index (dissimilarity index) for a set of compounds in R

Question

I want to calculate the diversity index for a given matrix.

I have a dataset matrix (xmatrix.RData), which is a 986 * 881 matrix, indicating 986 compounds and 881 fingerprints descriptors.

The formula to calculate the diversity index is explained in:
http://r.789695.n4.nabble.com/file/n4677766/Diversity_Index_Formula.pdf

score 2 · Accepted Answer · answered Oct 09 '13 at 16:13

2

I would do something like this:

# this calulates you the dissimilarity matrix diss(i, j) in the paper
# look at help page of dist for different methods like euclidean, maximum...
diss<-as.matrix(dist(xmatrix, method="binary", diag=TRUE, upper=TRUE))
l<-nrow(xmatrix)

# the overall dissimilarity
sum(rowSums(diss))/(l*(l-1))

hope this helps...

answered Oct 09 '13 at 16:13

holzben

1,459
16
24

Thanks holzben. It works. That was what I want. It will be better if the computational speed for "diss" is faster. – BioChemoinformatics Oct 09 '13 at 18:36

Calculate diversity index (dissimilarity index) for a set of compounds in R

1 Answers1