Minimum samples to cluster & cluster count linked about kepler-mapper HOT 6 OPEN

This is a bit confusing. If you have any ideas for how to fix it, a PR would be very welcome.

I think this started as a necessary condition when using a clustering method that requires n many samples or will fail and was not meant as a user facing facility.

Do you want to expose the option to set min_cluster_samples to the user in a better way?

from kepler-mapper.

Dear all,

I ran into the same odd interaction and decided to make explicit the number of cells. I also had some odd interaction with agglomerative clustering and min_cluster_samples.

I think that it would be better to expose the option min_cluster_samples to set to the user.

Kind regards,

from kepler-mapper.

Cool, we can do that.

Originally this was added, because some cluster algorithms in Scikit-learn exit with an error if you try clustering data that has a size less than n_clusters. (n_clusters=3 and 2 samples in hypercube results in an error).

from kepler-mapper.

Hi @rintrah, Pull requests are very welcome if you have already made some of these changes!

Thank you!

from kepler-mapper.

Hello, Sauln,

I am sorry for my late response, but the past days have been hectic.

I will write a clean version and do a pull request. I haven't done pull request before, but I think it won't be so difficult.

Best wishes,

from kepler-mapper.

No worries!

We look forward to the PR and will try to help in whatever way possible to make the experience easy.

Please let me know if you have any questions.

from kepler-mapper.