Matplotlib for Python Developers - Second Edition

Extracting falsely predicted images

Similar to what we did at the beginning of this chapter, we will draw the digit images out. This time, we pick out the mislabeled ones because they are the ones we're concerned about. We will again pick 10 images and put them in a grid of subplots. We write the true label in green at the bottom as xlabel and the false label predicted in red as the title at the top for each image in a subplot:

import matplotlib.pyplot as pltnrows, ncols = 2, 5plt.figure(figsize=(6,3))for i in range(ncols * nrows):    j = mislabeled_indices[i]    ax = plt.subplot(nrows, ncols, i + 1)    ax.imshow(X_test[j].reshape(8,8),cmap='gray_r')    plt.xticks([])    plt.yticks([])    plt.title(y_test1[j],color='red') plt.xlabel(y_test[j],color='green') ...