SciPy — Clustering

K-means clustering

from scipy.cluster.vq import kmeans, vq
import numpy as np

# Generate data
data = np.random.rand(100, 2)

# Cluster into 3 groups
centroids, distortion = kmeans(data, 3)
print(f"Centroids:\n{centroids}")
print(f"Distortion: {distortion:.4f}")

# Assign points to clusters
labels, _ = vq(data, centroids)
print(f"Labels: {np.unique(labels)}")

Hierarchical clustering

from scipy.cluster.hierarchy import linkage, fcluster
import numpy as np

data = np.random.rand(50, 2)

# Create linkage matrix
Z = linkage(data, method='ward')

# Form clusters
clusters = fcluster(Z, t=3, criterion='maxclust')
print(f"Clusters: {np.unique(clusters)}")

Dendrogram

from scipy.cluster.hierarchy import dendrogram
import matplotlib.pyplot as plt

dendrogram(Z)
plt.title('Dendrogram')
plt.show()

Distance matrices

from scipy.spatial.distance import pdist, squareform
import numpy as np

data = np.random.rand(10, 2)

# Compute pairwise distances
distances = pdist(data)
dist_matrix = squareform(distances)
print(f"Distance matrix shape: {dist_matrix.shape}")

Silhouette analysis

from sklearn.metrics import silhouette_score
from scipy.cluster.vq import kmeans, vq

data = np.random.rand(100, 2)
centroids, _ = kmeans(data, 3)
labels, _ = vq(data, centroids)

score = silhouette_score(data, labels)
print(f"Silhouette score: {score:.4f}")

Mini Practice

1. Perform K-means clustering
2. Create hierarchical clustering
3. Visualize dendrogram
4. Evaluate clustering quality

Up Next

Continue with I/O - File operations.