def __call__(self, x, y):
x, y = x.get(), y.get()
if x is y:
return B.fill_diag(B.one(x), B.shape(uprank(x))[0])
else:
x, y = uprank(x), uprank(y)
return Zero(B.dtype(x), B.shape(x)[0], B.shape(y)[0])
def __call__(self, x, y):
w_x, w_y = x.w, y.w
x, y = x.get(), y.get()
if x is y:
return Diagonal(1 / w_x)
else:
return Zero(B.dtype(x), num_elements(x), num_elements(y))
def __call__(self, x, y):
w_x, w_y = x.w, y.w
x, y = x.get(), y.get()
if x is y:
return Diagonal(1 / w_x)
else:
x, y = uprank(x), uprank(y)
return Zero(B.dtype(x), B.shape(x)[0], B.shape(y)[0])
def _noise_as_matrix(noise: type(None), dtype: B.DType, n: int):
"""Efficiently represent noise as a matrix.
Args:
noise (None, scalar, vector, or matrix): Noise. `None` means no noise.
dtype (dtype): Data type that the noise should be.
n (int): Number of observations.
Returns:
matrix: Noise as a matrix.
"""
return Zero(dtype, n, n)
def test_normal_mean_is_zero():
# Check zero case.
dist = Normal(B.eye(3))
assert dist.mean_is_zero
approx(dist.mean, B.zeros(3, 1))
# Check another zero case.
dist = Normal(Zero(np.float32, 3, 1), B.eye(3))
assert dist.mean_is_zero
approx(dist.mean, B.zeros(3, 1))
# Check nonzero case.
assert not Normal(B.randn(3, 1), B.eye(3)).mean_is_zero
def test_zero_attributes():
zero = Zero(int, 3, 4)
assert zero.dtype == int
assert zero.rows == 3
assert zero.cols == 4
def test_zero_formatting():
assert str(Zero(int, 3, 3)) == "<zero matrix: shape=3x3, dtype=int>"
assert repr(Zero(int, 3, 3)) == "<zero matrix: shape=3x3, dtype=int>"
def generate(code):
"""Generate a random tensor of a particular type, specified with a code.
Args:
code (str): Code of the matrix.
Returns:
tensor: Random tensor.
"""
mat_code, shape_code = code.split(":")
# Parse shape.
if shape_code == "":
shape = ()
else:
shape = tuple(int(d) for d in shape_code.split(","))
if mat_code == "randn":
return B.randn(*shape)
elif mat_code == "randn_pd":
mat = B.randn(*shape)
# If it is a scalar or vector, just pointwise square it.
if len(shape) in {0, 1}:
return mat**2 + 1
else:
return B.matmul(mat, mat, tr_b=True) + B.eye(shape[0])
elif mat_code == "zero":
return Zero(B.default_dtype, *shape)
elif mat_code == "const":
return Constant(B.randn(), *shape)
elif mat_code == "const_pd":
return Constant(B.randn()**2 + 1, *shape)
elif mat_code == "lt":
mat = B.vec_to_tril(B.randn(int(0.5 * shape[0] * (shape[0] + 1))))
return LowerTriangular(mat)
elif mat_code == "lt_pd":
mat = generate(f"randn_pd:{shape[0]},{shape[0]}")
return LowerTriangular(B.cholesky(B.reg(mat)))
elif mat_code == "ut":
mat = B.vec_to_tril(B.randn(int(0.5 * shape[0] * (shape[0] + 1))))
return UpperTriangular(B.transpose(mat))
elif mat_code == "ut_pd":
mat = generate(f"randn_pd:{shape[0]},{shape[0]}")
return UpperTriangular(B.transpose(B.cholesky(B.reg(mat))))
elif mat_code == "dense":
return Dense(generate(f"randn:{shape_code}"))
elif mat_code == "dense_pd":
return Dense(generate(f"randn_pd:{shape_code}"))
elif mat_code == "diag":
return Diagonal(generate(f"randn:{shape_code}"))
elif mat_code == "diag_pd":
return Diagonal(generate(f"randn_pd:{shape_code}"))
else:
raise RuntimeError(f'Cannot parse generation code "{code}".')
def elwise(self, x, y):
if x is y and B.shape(B.uprank(x))[0] == B.shape(self.noises)[0]:
return B.uprank(self.noises)
else:
x = B.uprank(x)
return Zero(B.dtype(x), B.shape(x)[0], 1)
def __call__(self, x, y):
if x is y and B.shape(uprank(x))[0] == B.shape(self.noises)[0]:
return Diagonal(self.noises)
else:
x, y = uprank(x), uprank(y)
return Zero(B.dtype(x), B.shape(x)[0], B.shape(y)[0])
def __call__(self, x, y):
return Zero(B.dtype(x), B.shape(x)[0], B.shape(y)[0])
def elwise(self, x, y):
if x is y and num_elements(x) == B.shape(self.noises)[0]:
return uprank(self.noises)
else:
return Zero(B.dtype(x), num_elements(x), 1)
def __call__(self, x, y):
if x is y and num_elements(x) == B.shape(self.noises)[0]:
return Diagonal(self.noises)
else:
return Zero(B.dtype(x), num_elements(x), num_elements(y))
def __call__(self, x, y):
y = y.get()
return Zero(B.dtype(x), num_elements(x), num_elements(y))
def test_isabstract_zero():
check_isabstract(lambda: Zero(np.float64, 2, 2))
def __call__(self, x, y):
x, y = uprank(x), uprank(y.get())
return Zero(B.dtype(x), B.shape(x)[0], B.shape(y)[0])
def __call__(self, x, y):
x, y = x.get(), y.get()
if x is y:
return B.fill_diag(B.one(x), num_elements(x))
else:
return Zero(B.dtype(x), num_elements(x), num_elements(y))