search

lachhebo / pyclustertend

A python package to assess cluster tendency
https://pyclustertend.readthedocs.io/en/master/index.html
BSD 3-Clause "New" or "Revised" License
46 stars 8 forks source link

Bug? #16

Open Muuuh219 opened 2 years ago

Muuuh219 commented 2 years ago

Hello,

First of all, thank you for implementiv VAT and iVAT! In the function compute_ivat_ordered_dissimilarity_matrix, the iVAT matrix is computed, but it is not symmetric in the end. Is is possible that there is some bug? I fixed it for now with the following code: I changed return re_ordered_matrix to for i in range(re_ordered_matrix.shape[0]): for j in range(i): re_ordered_matrix[j, i] = re_ordered_matrix[i, j] return re_ordered_matrix

Sincere regards!

lachhebo commented 2 years ago

Hello,

Thanks, can you tell me which version of the package you are using and if possible an example on which the ivat algorithm return a non symmetric matrix.

Thanks

Muuuh219 commented 2 years ago

Hello Mr Lachheb,Thank you very much for your answer. I do not know the version, but it should be the newest (from the time I asked the question) and it happens with any matrix. It can also be seen in your documentation, I attached a screenshot where the upper triangular matrix has a lot of black dots.Sincere regards,Nguyen -------- Ursprüngliche Nachricht --------Von: Ismaïl Lachheb @.> Datum: 31.08.22 00:21 (GMT+01:00) An: lachhebo/pyclustertend @.> Cc: Muuuh219 @.>, Author @.> Betreff: Re: [lachhebo/pyclustertend] Bug? (Issue #16) Hello, Thanks, can you tell me which version of the package you are using and if possible an example on which the ivat algorithm return a non symmetric matrix. Thanks

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: @.***>