test_kernel_approximation.py 文件源码

def test_rbf_sampler():
    # test that RBFSampler approximates kernel on random data
    # compute exact kernel
    gamma = 10.
    kernel = rbf_kernel(X, Y, gamma=gamma)

    # approximate kernel mapping
    rbf_transform = RBFSampler(gamma=gamma, n_components=1000, random_state=42)
    X_trans = rbf_transform.fit_transform(X)
    Y_trans = rbf_transform.transform(Y)
    kernel_approx = np.dot(X_trans, Y_trans.T)

    error = kernel - kernel_approx
    assert_less_equal(np.abs(np.mean(error)), 0.01)  # close to unbiased
    np.abs(error, out=error)
    assert_less_equal(np.max(error), 0.1)  # nothing too far off
    assert_less_equal(np.mean(error), 0.05)  # mean is fairly close