PCA-step3And4

lam, v = np.linalg.eig(cov)
print("Eigenvalues:")
print(lam)
print("Eigenvectors:")
print(v)
# Sort
eigvals, eigvects = (list(t) for t in zip(*sorted(zip(lam, v))))

# And now we can see what this looks like!
x_s = np.linspace(np.min(X_adj), np.max(X_adj), 10)
with plt.xkcd():
    plt.scatter(X_adj, Y_adj, label='Data', alpha=0.6)
    plt.plot(x_s, eigvects[0][0]/eigvects[0][1]*x_s, label='eigen 1')
    plt.plot(x_s, eigvects[1][0]/eigvects[1][1]*x_s, label='eigen 2')
    plt.legend()