def test_basic_arithmetic():
m1 = TensorProductMean(f1)
m2 = TensorProductMean(f2)
m3 = ZeroMean()
x1 = B.randn(10, 2)
x2 = B.randn()
approx((m1 * m2)(x1), m1(x1) * m2(x1))
approx((m1 * m2)(x2), m1(x2) * m2(x2))
approx((m1 + m3)(x1), m1(x1) + m3(x1))
approx((m1 + m3)(x2), m1(x2) + m3(x2))
approx((5.0 * m1)(x1), 5.0 * m1(x1))
approx((5.0 * m1)(x2), 5.0 * m1(x2))
approx((5.0 + m1)(x1), 5.0 + m1(x1))
approx((5.0 + m1)(x2), 5.0 + m1(x2))
def test_derivative():
m = TensorProductMean(lambda x: x**2)
m2 = TensorProductMean(lambda x: x**3)
x = B.randn(tf.float64, 10, 1)
allclose(m.diff(0)(x), 2 * x)
allclose(m2.diff(0)(x), 3 * x**2)
def test_posterior_mean():
z = B.linspace(0, 1, 10)
pcm = PosteriorMean(
TensorProductMean(lambda x: x),
TensorProductMean(lambda x: x ** 2),
EQ(),
z,
2 * EQ()(z),
B.randn(10),
)
# Check name.
assert str(pcm) == "PosteriorMean()"
# Check that the mean computes.
pcm(z)
def test_tensor_product():
m1 = 5 * OneMean() + (lambda x: x ** 2)
m2 = (lambda x: x ** 2) + 5 * OneMean()
m3 = (lambda x: x ** 2) + ZeroMean()
m4 = ZeroMean() + (lambda x: x ** 2)
x = B.randn(10, 1)
assert np.allclose(m1(x), 5 + x ** 2)
assert np.allclose(m2(x), 5 + x ** 2)
assert np.allclose(m3(x), x ** 2)
assert np.allclose(m4(x), x ** 2)
def my_function(x):
pass
assert str(TensorProductMean(my_function)) == "my_function"
def test_input_transform():
m = 5 * OneMean() + (lambda x: x ** 2)
x = B.randn(10, 3)
approx(m.transform(lambda x: x - 5)(x), m(x - 5))
def test_stretching():
m = 5 * OneMean() + (lambda x: x ** 2)
x = B.randn(10, 3)
approx(m.stretch(5)(x), m(x / 5))
def test_shifting():
m = 5 * OneMean() + (lambda x: x ** 2)
x = B.randn(10, 3)
approx(m.shift(5)(x), m(x - 5))
def test_selected_mean():
m = 5 * OneMean() + (lambda x: x ** 2)
x = B.randn(10, 3)
approx(m.select([1, 2])(x), m(x[:, [1, 2]]))
@_dispatch(Number)
def f2(x):
return np.array([[x ** 3]])
@_dispatch(object)
def f2(x):
return B.sum(x ** 3, axis=1)[:, None]
def test_corner_cases():
with pytest.raises(RuntimeError):
Mean()(1.0)
@pytest.mark.parametrize("x", [B.randn(10), Input(B.randn(10)), FDD(None, B.randn(10))])
def test_construction(x):
m = TensorProductMean(lambda y: y ** 2)
m(x)
# Test `MultiInput` construction.
approx(m(MultiInput(x, x)), B.concat(m(x), m(x), axis=0))
def test_basic_arithmetic():
m1 = TensorProductMean(f1)
m2 = TensorProductMean(f2)
m3 = ZeroMean()
x1 = B.randn(10, 2)
x2 = B.randn()