In [1]:

    

%matplotlib nbagg


    

In [2]:

    

from sklearn.datasets import make_blobs
import seaborn as sns
import hdbscan


    

In [3]:

    

X, Y = make_blobs(n_features=2, centers=[(1,1), (1,1.5), (3,2)],
                    n_samples=90,
                    random_state=42,
                    center_box=[0,1,0,1],
                    cluster_std=0.15,
                   )
plt.scatter(X[:, 0], X[:, 1], marker='o', c=Y)


    

    















    












    
Out[3]:





<matplotlib.collections.PathCollection at 0x111057da0>

In [4]:

    

import hdbscan
for i in range(3, 30):
    for j in range(1, i):
        clusterer = hdbscan.HDBSCAN(min_cluster_size=i, 
                                    min_samples=j,
                                    gen_min_span_tree=True,
                                   )

        clusterer.fit(X)
        n_found = clusterer.labels_.max() + 1
        if n_found > 2:
            print("Clusters found:", n_found)
            print("i:",i,'\nj:', j)


    

    




Clusters found: 11
i: 3 
j: 1