def test_conditioning(generate_noise_tuple):
m = Measure()
p1 = GP(EQ(), measure=m)
p2 = GP(Exp(), measure=m)
p_sum = p1 + p2
# Sample some data to condition on.
x1 = B.linspace(0, 2, 3)
n1 = generate_noise_tuple(x1)
y1 = p1(x1, *n1).sample()
tup1 = (p1(x1, *n1), y1)
x_sum = B.linspace(3, 5, 3)
n_sum = generate_noise_tuple(x_sum)
y_sum = p_sum(x_sum, *n_sum).sample()
tup_sum = (p_sum(x_sum, *n_sum), y_sum)
# Determine FDDs to check.
x_check = B.linspace(0, 5, 5)
fdds_check = [
cross(p1, p2, p_sum)(x_check),
p1(x_check),
p2(x_check),
p_sum(x_check),
]
assert_equal_measures(
fdds_check,
m.condition(*tup_sum),
m.condition(tup_sum),
m | tup_sum,
m | (tup_sum, ),
m | Obs(*tup_sum),
m | Obs(tup_sum),
)
assert_equal_measures(
fdds_check,
m.condition(tup1, tup_sum),
m | (tup1, tup_sum),
m | Obs(tup1, tup_sum),
)
# Check that conditioning gives an FDD and that it is consistent.
post = m | tup1
assert isinstance(post(p1(x1, 0.1)), FDD)
assert_equal_measures(post(p1(x1, 0.1)), post(p1)(x1, 0.1))
def test_pseudo_conditioning_and_elbo(generate_noise_tuple):
m = Measure()
p1 = GP(EQ(), measure=m)
p2 = GP(Exp(), measure=m)
p_sum = p1 + p2
# Sample some data to condition on.
x1 = B.linspace(0, 2, 3)
n1 = generate_noise_tuple(x1)
y1 = p1(x1, *n1).sample()
tup1 = (p1(x1, *n1), y1)
x_sum = B.linspace(3, 5, 3)
n_sum = generate_noise_tuple(x_sum)
y_sum = p_sum(x_sum, *n_sum).sample()
tup_sum = (p_sum(x_sum, *n_sum), y_sum)
# Determine FDDs to check.
x_check = B.linspace(0, 5, 5)
fdds_check = [
cross(p1, p2, p_sum)(x_check),
p1(x_check),
p2(x_check),
p_sum(x_check),
]
# Check conditioning and ELBO on one data set.
assert_equal_measures(
fdds_check,
m | tup_sum,
m | PseudoObs(p_sum(x_sum), *tup_sum),
m | PseudoObs((p_sum(x_sum), ), *tup_sum),
m | PseudoObs((p_sum(x_sum), p1(x1)), *tup_sum),
m | PseudoObs(p_sum(x_sum), tup_sum),
m | PseudoObs((p_sum(x_sum), ), tup_sum),
m.condition(PseudoObs((p_sum(x_sum), p1(x1)), tup_sum)),
)
approx(
m.logpdf(Obs(*tup_sum)),
PseudoObs(p_sum(x_sum), tup_sum).elbo(m),
)
# Check conditioning and ELBO on two data sets.
assert_equal_measures(
fdds_check,
m | (tup_sum, tup1),
m.condition(PseudoObs((p_sum(x_sum), p1(x1)), tup_sum, tup1)),
)
approx(
m.logpdf(Obs(tup_sum, tup1)),
PseudoObs((p_sum(x_sum), p1(x1)), tup_sum, tup1).elbo(m),
)
# The following lose information, so check them separately.
assert_equal_measures(
fdds_check,
m | PseudoObs(p_sum(x_sum), tup_sum, tup1),
m | PseudoObs((p_sum(x_sum), ), tup_sum, tup1),
)
# Test caching.
for name in ["K_z", "elbo", "mu", "A"]:
obs = PseudoObs(p_sum(x_sum), *tup_sum)
assert getattr(obs, name)(m) is getattr(obs, name)(m)
# Test requirement that noise must be diagonal.
with pytest.raises(RuntimeError):
PseudoObs(p_sum(x_sum), (p_sum(x_sum,
p_sum(x_sum).var), y_sum)).elbo(m)
# Test that noise on inducing points loses information.
with pytest.raises(AssertionError):
assert_equal_measures(
fdds_check,
m | tup_sum,
m | PseudoObs(p_sum(x_sum, 0.1), *tup_sum),
)