def test_default_dtype(self):
random = RandomStreams(utt.fetch_seed())
low = tensor.dscalar()
high = tensor.dscalar()
# Should not silently downcast from low and high
out0 = random.uniform(low=low, high=high, size=(42, ))
assert out0.dtype == "float64"
f0 = function([low, high], out0)
val0 = f0(-2.1, 3.1)
assert val0.dtype == "float64"
# Should downcast, since asked explicitly
out1 = random.uniform(low=low, high=high, size=(42, ), dtype="float32")
assert out1.dtype == "float32"
f1 = function([low, high], out1)
val1 = f1(-1.1, 1.1)
assert val1.dtype == "float32"
# Should use floatX
lowf = tensor.fscalar()
highf = tensor.fscalar()
outf = random.uniform(low=lowf, high=highf, size=(42, ))
assert outf.dtype == config.floatX
ff = function([lowf, highf], outf)
valf = ff(np.float32(-0.1), np.float32(0.3))
assert valf.dtype == config.floatX
def test_param_allow_downcast_floatX(self):
a = tensor.fscalar("a")
b = tensor.fscalar("b")
c = tensor.fscalar("c")
f = pfunc(
[
In(a, allow_downcast=True),
In(b, allow_downcast=False),
In(c, allow_downcast=None),
],
(a + b + c),
)
# If the values can be accurately represented, everything is OK
assert np.all(f(0, 0, 0) == 0)
# If allow_downcast is True, idem
assert np.allclose(f(0.1, 0, 0), 0.1)
# If allow_downcast is False, nope
with pytest.raises(TypeError):
f(0, 0.1, 0)
# If allow_downcast is None, it should work iff floatX=float32
if config.floatX == "float32":
assert np.allclose(f(0, 0, 0.1), 0.1)
else:
with pytest.raises(TypeError):
f(0, 0, 0.1)
def test_copy_delete_updates(self):
w = tt.iscalar("w")
x = tt.fscalar("x")
# SharedVariable for tests, one of them has update
y = aesara.shared(value=1, name="y")
z = aesara.shared(value=2, name="z")
out = x + y + z
# Test for different linkers
# for mode in ["FAST_RUN","FAST_COMPILE"]:
# second_time = False
for mode in ["FAST_RUN", "FAST_COMPILE"]:
ori = aesara.function([x], out, mode=mode, updates={z: z * 2})
cpy = ori.copy(delete_updates=True)
assert cpy(1)[0] == 4
assert cpy(1)[0] == 4
assert cpy(1)[0] == 4
# Test if unused implicit and explicit inputs from delete_updates
# are ignored as intended.
for mode in ["FAST_RUN", "FAST_COMPILE"]:
ori = aesara.function([x], x, mode=mode, updates={z: z * 2})
cpy = ori.copy(delete_updates=True)
ori = aesara.function([x, w], x, mode=mode, updates={z: z + w})
cpy = ori.copy(delete_updates=True)
def test_copy_share_memory(self):
x = tt.fscalar("x")
# SharedVariable for tests, one of them has update
y = aesara.shared(value=1)
z = aesara.shared(value=2)
out = tt.tanh((x + y + 2) / (x + z - 0.2) ** 2)
# Test for different linkers
for mode in ["FAST_RUN", "FAST_COMPILE"]:
ori = aesara.function([x], [out], mode=mode, updates={z: z + 1})
cpy = ori.copy(share_memory=True)
# Test if memories shared
storage_map_ori = ori.fn.storage_map
storage_map_cpy = cpy.fn.storage_map
fgraph_cpy = cpy.maker.fgraph
# Assert intermediate and Constants storages are shared.
# and output stoarges are not shared
i_o_variables = fgraph_cpy.inputs + fgraph_cpy.outputs
ori_storages = storage_map_ori.values()
l = [
val
for key, val in storage_map_cpy.items()
if key not in i_o_variables or isinstance(key, tt.Constant)
]
for storage in l:
assert any([storage is s for s in ori_storages])
# Assert storages of SharedVariable without updates are shared
for (input, _1, _2), here, there in zip(
ori.indices, ori.input_storage, cpy.input_storage
):
assert here.data is there.data
def test_allow_downcast_floatX(self):
a = tensor.fscalar("a")
b = tensor.fvector("b")
f = pfunc([a, b], (a + b), allow_input_downcast=True)
g = pfunc([a, b], (a + b), allow_input_downcast=False)
h = pfunc([a, b], (a + b), allow_input_downcast=None)
# If the values can be accurately represented, OK
assert np.all(f(0, [0]) == 0)
assert np.all(g(0, [0]) == 0)
assert np.all(h(0, [0]) == 0)
# For the vector: OK iff allow_input_downcast is True
assert np.allclose(f(0, [0.1]), 0.1)
with pytest.raises(TypeError):
g(0, [0.1])
with pytest.raises(TypeError):
h(0, [0.1])
# For the scalar: OK if allow_input_downcast is True,
# or None and floatX==float32
assert np.allclose(f(0.1, [0]), 0.1)
with pytest.raises(TypeError):
g(0.1, [0])
if config.floatX == "float32":
assert np.allclose(h(0.1, [0]), 0.1)
else:
with pytest.raises(TypeError):
h(0.1, [0])
def setup_method(self):
super().setup_method()
# Sample computation that involves tensors with different numbers
# of dimensions
self.input1 = tt.fmatrix()
self.input2 = tt.fscalar()
self.output = tt.dot((self.input1 - self.input2),
(self.input1 - self.input2).transpose())
# Declare the conditional breakpoint
self.breakpointOp = PdbBreakpoint("Sum of output too high")
self.condition = tt.gt(self.output.sum(), 1000)
(
self.monitored_input1,
self.monitored_input2,
self.monitored_output,
) = self.breakpointOp(self.condition, self.input1, self.input2,
self.output)