def test_gpujoin_gpualloc():
a = tt.fmatrix("a")
a_val = np.asarray(np.random.rand(4, 5), dtype="float32")
b = tt.fmatrix("b")
b_val = np.asarray(np.random.rand(3, 5), dtype="float32")
f = aesara.function(
[a, b], tt.join(0, tt.zeros_like(a), tt.ones_like(b)) + 4, mode=mode_without_gpu
)
f_gpu = aesara.function(
[a, b], tt.join(0, tt.zeros_like(a), tt.ones_like(b)), mode=mode_with_gpu
)
f_gpu2 = aesara.function(
[a, b], tt.join(0, tt.zeros_like(a), tt.ones_like(b)) + 4, mode=mode_with_gpu
)
assert sum([node.op == tt.alloc for node in f.maker.fgraph.toposort()]) == 2
assert sum([node.op == tt.join_ for node in f.maker.fgraph.toposort()]) == 1
assert (
sum([isinstance(node.op, GpuAlloc) for node in f_gpu.maker.fgraph.toposort()])
== 2
)
assert sum([node.op == gpu_join for node in f_gpu.maker.fgraph.toposort()]) == 1
assert (
sum([isinstance(node.op, GpuAlloc) for node in f_gpu2.maker.fgraph.toposort()])
== 2
)
assert sum([node.op == gpu_join for node in f_gpu2.maker.fgraph.toposort()]) == 1
assert np.allclose(f(a_val, b_val), f_gpu2(a_val, b_val))
def pad_dims(input, leftdims, rightdims):
"""Reshapes the input to a (leftdims + rightdims) tensor
This helper function is used to convert pooling inputs with arbitrary
non-pooling dimensions to the correct number of dimensions for the
GPU pooling ops.
This reduces or expands the number of dimensions of the input to
exactly `leftdims`, by adding extra dimensions on the left or by
combining some existing dimensions on the left of the input.
Use `unpad_dims` to reshape back to the original dimensions.
Examples
--------
Given input of shape (3, 5, 7), ``pad_dims(input, 2, 2)``
adds a singleton dimension and reshapes to (1, 3, 5, 7).
Given that output from pad_dims, ``unpad_dims(output, input, 2, 2)``
reshapes back to (3, 5, 7).
Given input of shape (3, 5, 7, 9), ``pad_dims(input, 2, 2)``
does not reshape and returns output with shape (3, 5, 7, 9).
Given input of shape (3, 5, 7, 9, 11), ``pad_dims(input, 2, 2)``
combines the first two dimensions and reshapes to (15, 7, 9, 11).
Given input of shape (3, 5, 7, 9), ``pad_dims(input, 2, 3)``
adds a singleton dimension and reshapes to (1, 3, 5, 7, 9).
"""
assert input.ndim >= rightdims
if input.ndim == (leftdims + rightdims):
return input
# extract image dimensions
img_shape = input.shape[-rightdims:]
non_pool_ndim = input.ndim - rightdims
if non_pool_ndim < leftdims:
# too few dimensions, pad on the left
dummy_dims = tensor.as_tensor([1] * (leftdims - non_pool_ndim))
new_shape = tensor.join(0, dummy_dims, input.shape[:non_pool_ndim], img_shape)
else:
# too many dimensions, combine the leading dimensions
batched_ndim = non_pool_ndim - leftdims + 1
batch_size = tensor.prod(input.shape[:batched_ndim])
# convert to a vector for tensor.join
batch_size = tensor.shape_padright(batch_size, 1)
new_shape = tensor.join(
0, batch_size, input.shape[batched_ndim:non_pool_ndim], img_shape
)
# store in the required shape
new_shape = tensor.cast(new_shape, "int64")
input_ND = GpuReshape(leftdims + rightdims)(input, new_shape)
return input_ND
def unpad_dims(output, input, leftdims, rightdims):
"""Reshapes the output after pad_dims.
This reverts the padding by `pad_dims`.
"""
if output.ndim == input.ndim:
return output
# restore the output to the original shape
outshp = tensor.join(0, input.shape[:-rightdims], output.shape[-rightdims:])
return GpuReshape(input.ndim)(output, outshp)
def normal(
self,
size,
avg=0.0,
std=1.0,
ndim=None,
dtype=None,
nstreams=None,
truncate=False,
**kwargs,
):
"""
Sample a tensor of values from a normal distribution.
Parameters
----------
size : int_vector_like
Array dimensions for the output tensor.
avg : float_like, optional
The mean value for the truncated normal to sample from (defaults to 0.0).
std : float_like, optional
The standard deviation for the truncated normal to sample from (defaults to 1.0).
truncate : bool, optional
Truncates the normal distribution at 2 standard deviations if True (defaults to False).
When this flag is set, the standard deviation of the result will be less than the one specified.
ndim : int, optional
The number of dimensions for the output tensor (defaults to None).
This argument is necessary if the size argument is ambiguous on the number of dimensions.
dtype : str, optional
The data-type for the output tensor. If not specified,
the dtype is inferred from avg and std, but it is at least as precise as floatX.
kwargs
Other keyword arguments for random number generation (see uniform).
Returns
-------
samples : TensorVariable
A Aesara tensor of samples randomly drawn from a normal distribution.
"""
size = _check_size(size)
avg = undefined_grad(as_tensor_variable(avg))
std = undefined_grad(as_tensor_variable(std))
if dtype is None:
dtype = aes.upcast(config.floatX, avg.dtype, std.dtype)
avg = at.cast(avg, dtype=dtype)
std = at.cast(std, dtype=dtype)
# generate even number of uniform samples
# Do manual constant folding to lower optiimizer work.
if isinstance(size, Constant):
n_odd_samples = size.prod(dtype="int64")
else:
n_odd_samples = prod(size, dtype="int64")
n_even_samples = n_odd_samples + n_odd_samples % 2
uniform = self.uniform(
(n_even_samples, ),
low=0.0,
high=1.0,
ndim=1,
dtype=dtype,
nstreams=nstreams,
**kwargs,
)
# box-muller transform
u1 = uniform[:n_even_samples // 2]
u2 = uniform[n_even_samples // 2:]
r = sqrt(-2.0 * log(u1))
theta = np.array(2.0 * np.pi, dtype=dtype) * u2
cos_theta, sin_theta = cos(theta), sin(theta)
z0 = r * cos_theta
z1 = r * sin_theta
if truncate:
# use valid samples
to_fix0 = (z0 < -2.0) | (z0 > 2.0)
to_fix1 = (z1 < -2.0) | (z1 > 2.0)
z0_valid = z0[at.nonzero(~to_fix0)]
z1_valid = z1[at.nonzero(~to_fix1)]
# re-sample invalid samples
to_fix0 = at.nonzero(to_fix0)[0]
to_fix1 = at.nonzero(to_fix1)[0]
n_fix_samples = to_fix0.size + to_fix1.size
lower = at.constant(1.0 / np.e**2, dtype=dtype)
u_fix = self.uniform(
(n_fix_samples, ),
low=lower,
high=1.0,
ndim=1,
dtype=dtype,
nstreams=nstreams,
**kwargs,
)
r_fix = sqrt(-2.0 * log(u_fix))
z0_fixed = r_fix[:to_fix0.size] * cos_theta[to_fix0]
z1_fixed = r_fix[to_fix0.size:] * sin_theta[to_fix1]
# pack everything together to a useful result
norm_samples = at.join(0, z0_valid, z0_fixed, z1_valid, z1_fixed)
else:
norm_samples = at.join(0, z0, z1)
if isinstance(n_odd_samples, Variable):
samples = norm_samples[:n_odd_samples]
elif n_odd_samples % 2 == 1:
samples = norm_samples[:-1]
else:
samples = norm_samples
samples = reshape(samples, newshape=size, ndim=ndim)
samples *= std
samples += avg
return samples
def conv2d(
input,
filters,
image_shape=None,
filter_shape=None,
border_mode="valid",
subsample=(1, 1),
**kargs,
):
"""
signal.conv.conv2d performs a basic 2D convolution of the input with the
given filters. The input parameter can be a single 2D image or a 3D tensor,
containing a set of images. Similarly, filters can be a single 2D filter or
a 3D tensor, corresponding to a set of 2D filters.
Shape parameters are optional and will result in faster execution.
Parameters
----------
input : Symbolic aesara tensor for images to be filtered.
Dimensions: ([num_images], image height, image width)
filters : Symbolic aesara tensor for convolution filter(s).
Dimensions: ([num_filters], filter height, filter width)
border_mode: {'valid', 'full'}
See scipy.signal.convolve2d.
subsample
Factor by which to subsample output.
image_shape : tuple of length 2 or 3
([num_images,] image height, image width).
filter_shape : tuple of length 2 or 3
([num_filters,] filter height, filter width).
kwargs
See aesara.tensor.nnet.conv.conv2d.
Returns
-------
symbolic 2D,3D or 4D tensor
Tensor of filtered images, with shape
([number images,] [number filters,] image height, image width).
"""
assert input.ndim in (2, 3)
assert filters.ndim in (2, 3)
# use shape information if it is given to us ###
if filter_shape and image_shape:
if input.ndim == 3:
bsize = image_shape[0]
else:
bsize = 1
imshp = (1, ) + tuple(image_shape[-2:])
if filters.ndim == 3:
nkern = filter_shape[0]
else:
nkern = 1
kshp = filter_shape[-2:]
else:
nkern, kshp = None, None
bsize, imshp = None, None
# reshape tensors to 4D, for compatibility with ConvOp ###
if input.ndim == 3:
sym_bsize = input.shape[0]
else:
sym_bsize = 1
if filters.ndim == 3:
sym_nkern = filters.shape[0]
else:
sym_nkern = 1
new_input_shape = aet.join(0, aet.stack([sym_bsize, 1]), input.shape[-2:])
input4D = reshape(input, new_input_shape, ndim=4)
new_filter_shape = aet.join(0, aet.stack([sym_nkern, 1]),
filters.shape[-2:])
filters4D = reshape(filters, new_filter_shape, ndim=4)
# perform actual convolution ###
op = conv.ConvOp(
output_mode=border_mode,
dx=subsample[0],
dy=subsample[1],
imshp=imshp,
kshp=kshp,
nkern=nkern,
bsize=bsize,
**kargs,
)
output = op(input4D, filters4D)
# flatten to 3D tensor if convolving with single filter or single image
if input.ndim == 2 and filters.ndim == 2:
if config.warn__signal_conv2d_interface:
warnings.warn(
"aesara.tensor.signal.conv2d() now outputs a 2d tensor when both"
" inputs are 2d. To disable this warning, set the Aesara flag"
" warn__signal_conv2d_interface to False",
stacklevel=3,
)
output = aet.flatten(output.T, ndim=2).T
elif input.ndim == 2 or filters.ndim == 2:
output = aet.flatten(output.T, ndim=3).T
return output
def test_join(self):
tv = np.asarray(self.rng.uniform(size=(10, )), aesara.config.floatX)
t = aesara.shared(tv)
out = aet.join(0, self.x, t)
self.check_rop_lop(out, (self.in_shape[0] + 10, ))