Skip to content

P
pytensor
提交 1294ff93 authored 作者: Nicolas Bouchard's avatar Nicolas Bouchard 提交者： Frederic
浏览文件
操作

Made corrections, add missing methods and tests.

上级 2edb35bc
隐藏空白字符变更
内嵌 并排
正在显示 包含 111 行增加 和 12 行删除
theano/sparse/sandbox/truedot.py
浏览文件 @ 1294ff93
import unittest import unittest
import theano
import numpy import numpy
import scipy.sparse as sp
from theano import sparse
from theano import gof, tensor, compile from theano import gof, tensor, compile
from theano.sparse.tests.test_basic import eval_outputs from theano.sparse.tests.test_basic import eval_outputs
from theano.sparse.basic import _( from theano.sparse.basic import (
is_sparse_variable, _is_dense_variable, _is_sparse_variable, _is_dense_variable,
as_sparse_variable, _is_sparse, _mtypes, _mtype_to_str) as_sparse_variable, _is_sparse, _mtypes, _mtype_to_str)
from theano.sparse import SparseType, dense_from_sparse, transpose from theano.sparse import SparseType, dense_from_sparse, transpose
from theano.sparse.tests.test_basic import sparse_random_inputs
from theano.tests import unittest_tools as utt
from theano.sparse import verify_grad_sparse
class TrueDot(gof.op.Op): class TrueDot(gof.op.Op):
"""Calculate the true dot operation between two matrices. """Calculate the true dot operation between two matrices.
...@@ -29,7 +35,7 @@ class TrueDot(gof.op.Op): ...@@ -29,7 +35,7 @@ class TrueDot(gof.op.Op):
:param x: Sparse matrix for the left operand. :param x: Sparse matrix for the left operand.
:param y: Sparse or dense matrix for the right operand. :param y: Sparse or dense matrix for the right operand.
:return: The dot product `x` . `y`. :return: The dot product `x` . `y` in a sparse matrix.
:note: :note:
- The grad implemented is regular, i.e. not structured. - The grad implemented is regular, i.e. not structured.
...@@ -46,7 +52,7 @@ class TrueDot(gof.op.Op): ...@@ -46,7 +52,7 @@ class TrueDot(gof.op.Op):
self.grad_preserves_dense == other.grad_preserves_dense) self.grad_preserves_dense == other.grad_preserves_dense)
def __hash__(self): def __hash__(self):
return hash(self.grad_preserves_dense) return hash(type(self)) ^ hash(self.grad_preserves_dense)
def __ne__(self, other): def __ne__(self, other):
return not (self == other) return not (self == other)
...@@ -71,7 +77,7 @@ class TrueDot(gof.op.Op): ...@@ -71,7 +77,7 @@ class TrueDot(gof.op.Op):
else: else:
raise NotImplementedError() raise NotImplementedError()
inputs = [x, y] # Need to convert? e.g. assparse inputs = [x, y] # Need to convert? e.g. assparse
outputs = [SparseType(dtype=x.type.dtype, outputs = [SparseType(dtype=x.type.dtype,
format=myformat).make_variable()] format=myformat).make_variable()]
return gof.Apply(self, inputs, outputs) return gof.Apply(self, inputs, outputs)
...@@ -86,19 +92,26 @@ class TrueDot(gof.op.Op): ...@@ -86,19 +92,26 @@ class TrueDot(gof.op.Op):
x, y = inp x, y = inp
out, = out_ out, = out_
rval = x.dot(y) rval = x.dot(y)
if not sp.issparse(rval):
rval = getattr(sp, x.format + '_matrix')(rval)
out[0] = rval out[0] = rval
def grad(self, inp, grads): def grad(self, (x, y), (gz, )):
x, y = inp
gz, = grads
assert _is_sparse_variable(gz) assert _is_sparse_variable(gz)
assert _is_sparse_variable(x) assert _is_sparse_variable(x)
rval = [true_dot(gz, y.T), true_dot(x.T, gz)] rval = [true_dot(gz, y.T), true_dot(x.T, gz)]
if _is_dense_variable(y): if _is_dense_variable(y):
if self.grad_preserves_dense: if self.grad_preserves_dense:
rval[1] = dense_from_sparse(rval[1]) rval[1] = dense_from_sparse(rval[1])
return rval return rval
def infer_shape(self, node, shapes):
return [(shapes[0][0], shapes[1][1])]
def __str__(self):
return self.__class__.__name__
def true_dot(x, y, grad_preserves_dense=True): def true_dot(x, y, grad_preserves_dense=True):
# TODO # TODO
...@@ -123,7 +136,93 @@ def true_dot(x, y, grad_preserves_dense=True): ...@@ -123,7 +136,93 @@ def true_dot(x, y, grad_preserves_dense=True):
return transpose(TrueDot(grad_preserves_dense)(y.T, x.T)) return transpose(TrueDot(grad_preserves_dense)(y.T, x.T))
class test_true_dot(unittest.TestCase): class TrueDotTester(utt.InferShapeTester):
def setUp(self):
super(TrueDotTester, self).setUp()
self.op = true_dot
self.op_class = TrueDot
def test_op_ss(self):
for format in sparse.sparse_formats:
for dtype in sparse.all_dtypes:
variable, data = sparse_random_inputs(format,
shape=(10, 10),
out_dtype=dtype,
n=2,
p=0.1)
f = theano.function(variable, self.op(*variable))
tested = f(*data)
x, y = [m.toarray() for m in data]
expected = numpy.dot(x, y)
assert tested.format == format
assert tested.dtype == expected.dtype
tested = tested.toarray()
assert numpy.allclose(tested, expected)
def test_op_sd(self):
for format in sparse.sparse_formats:
for dtype in sparse.all_dtypes:
variable, data = sparse_random_inputs(format,
shape=(10, 10),
out_dtype=dtype,
n=2,
p=0.1)
variable[1] = tensor.TensorType(dtype=dtype,
broadcastable=(False, False))()
data[1] = data[1].toarray()
f = theano.function(variable, self.op(*variable))
tested = f(*data)
expected = numpy.dot(data[0].toarray(), data[1])
assert tested.format == format
assert tested.dtype == expected.dtype
tested = tested.toarray()
assert numpy.allclose(tested, expected)
def test_infer_shape(self):
for format in sparse.sparse_formats:
for dtype in sparse.all_dtypes:
(x, ), (x_value, ) = sparse_random_inputs(format,
shape=(9, 10),
out_dtype=dtype,
p=0.1)
(y, ), (y_value, ) = sparse_random_inputs(format,
shape=(10, 24),
out_dtype=dtype,
p=0.1)
variable = [x, y]
data = [x_value, y_value]
self._compile_and_check(variable,
[self.op(*variable)],
data,
self.op_class)
def test_grad(self):
for format in sparse.sparse_formats:
for dtype in sparse.float_dtypes:
(x, ), (x_value, ) = sparse_random_inputs(format,
shape=(9, 10),
out_dtype=dtype,
p=0.1)
(y, ), (y_value, ) = sparse_random_inputs(format,
shape=(10, 24),
out_dtype=dtype,
p=0.1)
variable = [x, y]
data = [x_value, y_value]
verify_grad_sparse(
self.op,
data,
structured=False)
class TrueDotTester2(unittest.TestCase):
def setUp(self): def setUp(self):
numpy.random.seed(44) numpy.random.seed(44)
...@@ -210,7 +309,7 @@ class test_true_dot(unittest.TestCase): ...@@ -210,7 +309,7 @@ class test_true_dot(unittest.TestCase):
z = eval_outputs([zop]) z = eval_outputs([zop])
self.assertTrue(_is_sparse(z)) self.assertTrue(_is_sparse(z))
self.assertTrue(z.shape == (500, 2)) self.assertTrue(z.shape == (500, 2))
# self.assertTrue(type(z) is mtype) # self.assertTrue(type(z) is mtype)
w = mtype((500, 2)) w = mtype((500, 2))
w[(10, 0)] = 3. w[(10, 0)] = 3.
...@@ -254,7 +353,7 @@ class test_true_dot(unittest.TestCase): ...@@ -254,7 +353,7 @@ class test_true_dot(unittest.TestCase):
for epoch in xrange(50): for epoch in xrange(50):
y, loss, gw = trainfn(x, w) y, loss, gw = trainfn(x, w)
w = w - (lr * gw) w = w - (lr * gw)
print loss # print loss
self.assertTrue(origloss > loss) self.assertTrue(origloss > loss)
self.assertTrue('1.05191241115' == str(loss)) self.assertTrue('1.05191241115' == str(loss))
...@@ -265,7 +364,7 @@ class test_true_dot(unittest.TestCase): ...@@ -265,7 +364,7 @@ class test_true_dot(unittest.TestCase):
for mtype in _mtypes: for mtype in _mtypes:
x = tensor.matrix('x') x = tensor.matrix('x')
w = SparseType(dtype='float64', w = SparseType(dtype='float64',
format= mtype_to_str[mtype]).make_variable() format=_mtype_to_str[mtype]).make_variable()
xw = dense_from_sparse(true_dot(w, x)) xw = dense_from_sparse(true_dot(w, x))
y = dense_from_sparse(true_dot(w.T, xw)) y = dense_from_sparse(true_dot(w.T, xw))
diff = x - y diff = x - y
......
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论