I have clustered large dataset(~1m observations and ~200 features) with skmeans and I want to validate the results. The problem is that according to tutorial, skmeans produces object which is suitable for silhouette calculations, but if I follow the tutorial I get error message.Could somebody explain where does this error come from?Both objects are of the same class but plotting works for hparty,but not on my clustering.
class(hparty)
[1] "skmeans" "pclust"
hparty
A hard spherical k-means partition of 1504 objects into 5 classes.
Class sizes: 311, 278, 131, 429, 355
Call: skmeans(x = x, k = 5)
class(seg.sk10)
[1] "skmeans" "pclust"
seg.sk10
A hard spherical k-means partition of 1488441 objects into 10 classes.
Class sizes: 79007, 81885, 219730, 58998, 107383, 161908, 98570, 132689, 395367, 152904
Call: skmeans(x = duom.1, k = k)
plot(silhouette(seg.sk10))
Error in FUN(left, right) : non-numeric argument to binary operator
