Python Machine Learning By Example - Second Edition

Implementing k-means from scratch

We use the iris dataset from scikit-learn as an example. Let's first load the data and visualize it. We herein only use two features out of the original four for simplicity:

>>> from sklearn import datasets>>> iris = datasets.load_iris()>>> X = iris.data[:, 2:4]>>> y = iris.target

Since the dataset contains three iris classes, we plot it in three different colors, as follows:

>>> import numpy as np>>> from matplotlib import pyplot as plt>>> y_0 = np.where(y==0)>>> plt.scatter(X[y_0, 0], X[y_0, 1])>>> y_1 = np.where(y==1)>>> plt.scatter(X[y_1, 0], X[y_1, 1])>>> y_2 = np.where(y==2)>>> plt.scatter(X[y_2, 0], X[y_2, 1])>>> plt.show()

This will give you the following output for the origin data plot:

Assuming ...