Typically when I think of limiting the number of points in a cluster I
think of KD trees. I suppose that wouldn't work?
On Tue, Jul 11, 2017 at 11:22 AM, Ariani A wrote:
> ِDear Uri,
> Thanks. I just have a pairwise distance matrix and I want to implement it
> so that each cluster has at least 4
ِDear Uri,
Thanks. I just have a pairwise distance matrix and I want to implement it
so that each cluster has at least 40 data points. (in Agglomerative).
Does it work?
Thanks,
-Ariani
On Tue, Jul 11, 2017 at 1:54 PM, Uri Goren wrote:
> Take a look at scipy's fcluster function.
> If M is a matri
Take a look at scipy's fcluster function.
If M is a matrix of all of your feature vectors, this code snippet should
work.
You need to figure out what metric and algorithm work for you
from sklearn.metrics import pairwise_distance
from scipy.cluster import hierarchy
X = pairwise_dista
