scikit-learn Cookbook , Second Edition - Second Edition

How to do it...

1. Run all the transforms on iris_X. One of the transforms is t-SNE:

iris_pca = PCA(n_components = 2).fit_transform(iris_X)iris_tsne = TSNE(learning_rate=200).fit_transform(iris_X)iris_MDS = MDS(n_components = 2).fit_transform(iris_X)iris_ISO = Isomap(n_components = 2).fit_transform(iris_X)

2. Plot the results:

plt.figure(figsize=(20, 10))plt.subplot(221)plt.title('PCA')plt.scatter(iris_pca [:, 0], iris_pca [:, 1], c=y)plt.subplot(222)plt.scatter(iris_tsne[:, 0], iris_tsne[:, 1], c=y)plt.title('TSNE')plt.subplot(223)plt.scatter(iris_MDS[:, 0], iris_MDS[:, 1], c=y)plt.title('MDS')plt.subplot(224)plt.scatter(iris_ISO[:, 0], iris_ISO[:, 1], c=y)plt.title('ISO')

The t-SNE algorithm has been popular recently, yet it takes a lot ...