search

antklen / diameter-clustering

Clustering with maximum diameter (maximum distance between points inside clusters).
MIT License
24 stars 0 forks source link

label:enhancement will you parallelize the algorithm? #3

Open K-Mike opened 3 years ago

K-Mike commented 3 years ago

will you parallelize the algorithm?

antklen commented 3 years ago

Any ideas how to do it? =) First step is computing all pairwise distances between points, here I just use corresponding functions from sklearn, scipy, numpy. Second step is greedy. We add points one by one, comparing new points with all previously added. So it's not obvious how to parallelize it.