It includes non-core points, but not points that are out of eps from any core point. You can modify eps and min_samples. But perhaps you should just choose a different clustering algorithm if this is behaviour you absolutely do not want. On 30 January 2018 at 23:24, AMIR SHANEHSAZZADEH < amir.p.shanehsazzadeh@umasd.net> wrote:
Hello,
I am working with the latest implementation of DBSCAN. I believe that scikit-learn's implementation does not include non-core points in clusters. This results in border points not being included in clusters. Is there any way to remedy this issue so that border points are included in their respective clusters? Do you know what modifications I would need to make the code?
Thank you, Amir Shanehsazzadeh
_______________________________________________ scikit-learn mailing list scikit-learn@python.org https://mail.python.org/mailman/listinfo/scikit-learn