自定义数据集,使用scikit-learn 中K均值包 进行聚类

import matplotlib.pyplot as plt
from sklearn.cluster import KMeans
import numpy as np

class1_points = np.array([[1.9, 1.2],
                          [1.5, 2.1],
                          [1.9, 0.5],
                          [1.5, 0.9],
                          [0.9, 1.2],
                          [1.1, 1.7],
                          [1.4, 1.1]])

class2_points = np.array([[-1.9, 1.2],
                          [-1.5, 2.1],
                          [-1.9, 0.5],
                          [-1.5, 0.9],
                          [-0.9, 1.2],
                          [-1.1, 1.7],
                          [-1.4, 1.1]])

class3_points = np.array([[1.9, -1.2],
                          [1.5, -2.1],
                          [1.9, -0.5],
                          [1.5, -0.9],
                          [0.9, -1.2],
                          [1.1, -1.7],
                          [1.4, -1.1]])

class4_points = np.array([[-1.9, -1.2],
                          [-1.5, -2.1],
                          [-1.9, -0.5],
                          [-1.5, -0.9],
                          [-0.9, -1.2],
                          [-1.1, -1.7],
                          [-1.4, -1.1]])

data = np.concatenate((class1_points, class2_points, class3_points, class4_points))

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 6))

k = 4

ax1.scatter(data[:, 0], data[:, 1], s=50)
ax1.plot()

km = KMeans(n_clusters=k,max_iter=30)
km.fit(data)
centers = km.cluster_centers_
y_kmeans = km.predict(data)
print(y_kmeans)

for i in range(k):
    cluster_points = data[y_kmeans == i]
    centroid = centers[i]
    for cluster_points in cluster_points:
        ax2.plot([cluster_points[0], centroid[0]],[cluster_points[1], centroid[1]],'k--')

ax2.scatter(data[:, 0], data[:, 1], c = y_kmeans, s=50)
ax2.scatter(centers[:, 0], centers[:, 1], c='black', s=100, alpha=0.5)

plt.show()

自定义数据集,使用scikit-learn 中K均值包 进行聚类_第1张图片

你可能感兴趣的:(scikit-learn,均值算法)