hclust_yeast_demo.ipynb

Kernel: Unknown Kernel

import scipy.io
import os
from scipy.spatial.distance import pdist
try:
    from sklearn import cluster
except ModuleNotFoundError:
    %pip install -qq scikit-learn
    from sklearn import cluster
from sklearn.cluster import AgglomerativeClustering
from scipy.cluster.hierarchy import fcluster
import matplotlib.pyplot as plt
import numpy as np
import seaborn
try:
    import probml_utils as pml
except ModuleNotFoundError:
    %pip install -qq git+https://github.com/probml/probml-utils.git
    import probml_utils as pml

import requests
from io import BytesIO

url = 'https://github.com/probml/probml-data/blob/main/data/yeastData310.mat?raw=true'
response = requests.get(url)
rawdata = BytesIO(response.content)
data = scipy.io.loadmat(rawdata)

X = data['X']

corrDist = pdist(X, 'correlation')
clusterTree = AgglomerativeClustering(n_clusters=16, linkage="average")
clusterTree.fit(X)
clusters = clusterTree.labels_

fig, axes = plt.subplots(4, 4)
fig.suptitle('Hierarchical Clustering of Profiles ', y = 1 ,fontsize = 20)
times = data['times'].reshape(7, )
for c in range(0, 16):
    occurences = np.argwhere(clusters == (c))
    row = c//4
    col = c%4
    for occ in occurences:
        axes[row][col].plot(times, X[occ, :].reshape(7,))
    
plt.tight_layout(rect=[0, 0, 1, 0.90])
pml.savefig('clusterYeastHier16.pdf')
plt.show()

#plt.figure()
clusterYeastRowPerm = seaborn.clustermap(X[:, 1:])
#plt.title('hierarchical clustering')
pml.savefig('clusterYeastRowPerm.pdf')
plt.show()

plt.figure()
Z = scipy.cluster.hierarchy.linkage(corrDist, 'average')
scipy.cluster.hierarchy.dendrogram(Z, truncate_mode='lastp')
plt.title('average link')
plt.tick_params(labelbottom=False)
pml.savefig('clusterYeastAvgLink.pdf')
plt.show()

plt.figure()
Z = scipy.cluster.hierarchy.linkage(corrDist, 'complete')
scipy.cluster.hierarchy.dendrogram(Z, truncate_mode='lastp')
plt.title('complete link')
plt.tick_params(labelbottom=False)
pml.savefig('clusterYeastCompleteLink.pdf')
plt.show()

plt.figure()
Z = scipy.cluster.hierarchy.linkage(corrDist, 'single')
scipy.cluster.hierarchy.dendrogram(Z, truncate_mode='lastp')
plt.title('single link')
plt.tick_params(labelbottom=False)
pml.savefig('clusterYeastSingleLink.pdf')
plt.show()