Remove unique_ids line which causes problems and is not used

[Nuanda]

I think SimilarityClustering.R#prep line:

unique_ids <- makeUnique(sdfid(sdfset))

is not used at the moment, right? If it is so, I'd propose to remove it since it causes problems with some SDF libraries:

Error in seq.default(1, dupids[i]) : 'to' cannot be NA, NaN or infinite

This is probably connected to the fact that these libraries do not have anything in the molecule definition's 1st header line (blank molecule id).

[wjurkowski]

Yes, maybe we could have the redundancy as part of QC health check after user library upload

[Nuanda]

I'm not sure you you need to force uniqueness in such a way. It look like the method will apply its own unique cids (CMP1, CNP2...) anyway. Or, as I do, one can reassign cid with one's own unique ID set.

Can you remove that line?