ML - Lab 8 - Hierarchical Clustering: Dendrograms

import pandas as pd
import matplotlib.pyplot as plt
import sklearn

# Read and plot the data
X = [2, 8, 0, 7, 6]
Y = [0, 4, 6, 2, 1]
labels = ["x"+str(i) for i in range(1, len(X)+1)]
hdata = pd.DataFrame({"X": X, "Y": Y}, index=labels)
plt.figure()
plt.scatter(hdata.X, hdata.Y)
for i in range(len(hdata.index)):
    plt.text(hdata.loc[labels[i], "X"], hdata.loc[labels[i], "Y"], '%s' % (str(labels[i])), size=15, zorder=1)
plt.title("2-D Points")
plt.xlabel("X")
plt.ylabel("Y")
plt.show()


# Hierarchical Clustering
from scipy.cluster.hierarchy import dendrogram
from sklearn.cluster import AgglomerativeClustering
import numpy as np



# Single
linkage = "single"
n_clusters = None
clustering = AgglomerativeClustering(n_clusters=n_clusters, linkage=linkage, distance_threshold=0).fit(hdata)
linkage_matrix = np.column_stack([clustering.children_, clustering.distances_, np.ones(len(hdata.index)-1)]).astype(float)
plt.title("Linkage = " + linkage)
print("Linkage Matrix: ")
print(linkage_matrix)
print()
dendrogram(linkage_matrix, labels=labels)
plt.show()

# Complete
linkage = "complete"
n_clusters = None
clustering = AgglomerativeClustering(n_clusters=n_clusters, linkage=linkage, distance_threshold=0).fit(hdata)
linkage_matrix = np.column_stack([clustering.children_, clustering.distances_, np.ones(len(hdata.index)-1)]).astype(float)
plt.title("Linkage = " + linkage)
print("Linkage Matrix: ")
print(linkage_matrix)
print()
dendrogram(linkage_matrix, labels=labels)
plt.show()



# Single with 2 clusters - Draw with Different Colors
linkage = "single"
n_clusters = 2
clustering = AgglomerativeClustering(n_clusters=n_clusters, linkage=linkage).fit(hdata)
plt.scatter(hdata.X, hdata.Y, c=clustering.labels_, cmap="bwr")
for i in range(len(hdata.index)):
    plt.text(hdata.loc[labels[i], "X"], hdata.loc[labels[i], "Y"], '%s' % (str(labels[i])), size=15, zorder=1)
plt.title(linkage + " with " + str(n_clusters) + " clusters")
plt.show()

# Complete with 2 clusters - Draw with Different Colors
linkage = "complete"
n_clusters = 2
clustering = AgglomerativeClustering(n_clusters=n_clusters, linkage=linkage).fit(hdata)
plt.scatter(hdata.X, hdata.Y, c=clustering.labels_, cmap="bwr")
for i in range(len(hdata.index)):
    plt.text(hdata.loc[labels[i], "X"], hdata.loc[labels[i], "Y"], '%s' % (str(labels[i])), size=15, zorder=1)
plt.title(linkage + " with " + str(n_clusters) + " clusters")
plt.show()