I have a dissimilarity matrix computed by other R package, TSclust with INT.PER method. I want to use the dissimilarity matrix to retrieve optimal number of clusters from NbClus
Some of the indexes require numerical data.
Therefore, this function cannot be used on your data set, unless you disable these methods/indexes that require coordinates.
I'm trying to use the NbClust package as well. With a dissimilarity matrix, I was only able to run these 5 indexes:
indexes <- c('frey', 'dunn', 'cindex', 'silhouette', 'mcclain')