513 lines
18 KiB
Python
513 lines
18 KiB
Python
# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
import unittest
|
|
|
|
import numpy as np
|
|
from op_test import OpTest, get_places
|
|
|
|
import paddle
|
|
|
|
paddle.enable_static()
|
|
|
|
|
|
class TestDeterminantOp(OpTest):
|
|
def setUp(self):
|
|
self.python_api = paddle.linalg.det
|
|
self.init_data()
|
|
self.op_type = "determinant"
|
|
self.outputs = {'Out': self.target}
|
|
|
|
def test_check_output(self):
|
|
self.check_output(check_pir=True)
|
|
|
|
def test_check_grad(self):
|
|
self.check_grad(['Input'], ['Out'], check_pir=True)
|
|
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(3, 3, 3, 5, 5).astype('float64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase1(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(10, 10).astype('float32')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase1FP16(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(10, 10).astype(np.float16)
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case.astype(np.float32))
|
|
|
|
|
|
class TestDeterminantOpCase2(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
# not invertible matrix
|
|
self.case = np.ones([4, 2, 4, 4]).astype('float64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase2FP16(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
# not invertible matrix
|
|
self.case = np.ones([4, 2, 4, 4]).astype(np.float16)
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case.astype(np.float32)).astype(
|
|
np.float16
|
|
)
|
|
|
|
|
|
class TestDeterminantOpCase3(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.vectorize(complex)(
|
|
np.random.rand(10, 10), np.random.rand(10, 10)
|
|
).astype('complex64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase4(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.vectorize(complex)(
|
|
np.random.rand(10, 10), np.random.rand(10, 10)
|
|
).astype('complex128')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase5(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
# not invertible matrix
|
|
self.case = np.ones([4, 2, 4, 4]).astype('complex64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase6(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
# not invertible matrix
|
|
self.case = np.ones([4, 2, 4, 4]).astype('complex128')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase7(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.vectorize(complex)(
|
|
np.random.rand(5, 3, 10, 10), np.random.rand(5, 3, 10, 10)
|
|
).astype('complex64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOpCase8(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.vectorize(complex)(
|
|
np.random.rand(5, 3, 10, 10), np.random.rand(5, 3, 10, 10)
|
|
).astype('complex128')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOp_ZeroSize(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(0, 10, 10)
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantOp_ZeroSize2(TestDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(0, 0, 0)
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.linalg.det(self.case)
|
|
|
|
|
|
class TestDeterminantAPI(unittest.TestCase):
|
|
def setUp(self):
|
|
np.random.seed(0)
|
|
self.dtype = np.float32
|
|
self.shape = [3, 3, 5, 5]
|
|
self.x = np.random.random(self.shape).astype(self.dtype)
|
|
self.place = paddle.CPUPlace()
|
|
|
|
def test_api_static(self):
|
|
paddle.enable_static()
|
|
with paddle.static.program_guard(paddle.static.Program()):
|
|
x = paddle.static.data('X', self.shape, dtype=self.dtype)
|
|
out_value = paddle.linalg.det(x)
|
|
exe = paddle.static.Executor(self.place)
|
|
(out_np,) = exe.run(feed={'X': self.x}, fetch_list=[out_value])
|
|
out_ref = np.linalg.det(self.x)
|
|
|
|
np.testing.assert_allclose(out_np, out_ref, rtol=0.001)
|
|
self.assertEqual(out_np.shape, out_ref.shape)
|
|
self.assertEqual(tuple(out_value.shape), out_ref.shape)
|
|
|
|
def test_api_dygraph(self):
|
|
paddle.disable_static(self.place)
|
|
x_tensor = paddle.to_tensor(self.x)
|
|
out = paddle.linalg.det(x_tensor)
|
|
out_ref = np.linalg.det(self.x)
|
|
np.testing.assert_allclose(out.numpy(), out_ref, rtol=0.001)
|
|
paddle.enable_static()
|
|
|
|
|
|
def determinant_complex_numeric_grad_single_batch(
|
|
x, n, delta=0.005, det_out_grad=np.array(1 + 0j)
|
|
):
|
|
# an naive implementation of numeric_grad with single batch input x
|
|
# the output of det for complex matrix is always complex, so det_out_grad
|
|
# should be a+bj, where a and b are arbitrary real numbers
|
|
dx = []
|
|
for i in range(n):
|
|
for j in range(n):
|
|
xp = x.copy()
|
|
xn = x.copy()
|
|
xpj = x.copy()
|
|
xnj = x.copy()
|
|
xp[i, j] += delta
|
|
xn[i, j] -= delta
|
|
xpj[i, j] += delta * 1j
|
|
xnj[i, j] -= delta * 1j
|
|
yp = np.linalg.det(xp)
|
|
yn = np.linalg.det(xn)
|
|
ypj = np.linalg.det(xpj)
|
|
ynj = np.linalg.det(xnj)
|
|
df_over_dr = (yp - yn) / delta / 2
|
|
df_over_di = (ypj - ynj) / delta / 2
|
|
dl_over_du, dl_over_dv = det_out_grad.real, det_out_grad.imag
|
|
du_over_dr, dv_over_dr = df_over_dr.real, df_over_dr.imag
|
|
du_over_di, dv_over_di = df_over_di.real, df_over_di.imag
|
|
dl_over_dr = np.sum(
|
|
dl_over_du * du_over_dr + dl_over_dv * dv_over_dr
|
|
)
|
|
dl_over_di = np.sum(
|
|
dl_over_du * du_over_di + dl_over_dv * dv_over_di
|
|
)
|
|
dx.append(dl_over_dr + 1j * dl_over_di)
|
|
return np.array(dx).reshape([n, n])
|
|
|
|
|
|
class TestDeterminantAPIComplex(unittest.TestCase):
|
|
def setUp(self):
|
|
np.random.seed(0)
|
|
self.dtype = np.complex64
|
|
self.shape = [2, 1, 4, 3, 6, 6]
|
|
self.x = np.vectorize(complex)(
|
|
np.random.random(self.shape), np.random.random(self.shape)
|
|
).astype(self.dtype)
|
|
self.place = paddle.CPUPlace()
|
|
self.out_grad = (
|
|
np.array([1 - 0.5j] * 2 * 1 * 4 * 3)
|
|
.reshape(2, 1, 4, 3)
|
|
.astype(self.dtype)
|
|
)
|
|
self.x_grad_ref_dy = self.get_numeric_grad(
|
|
self.x, self.shape, self.out_grad
|
|
)
|
|
self.x_grad_ref_st = self.get_numeric_grad(self.x, self.shape)
|
|
|
|
def get_numeric_grad(self, x, shape, out_grad=None):
|
|
n = shape[-1]
|
|
flatten_x = x.reshape([-1, n, n])
|
|
n_batch = flatten_x.shape[0]
|
|
grad = []
|
|
if out_grad is None:
|
|
for b in range(n_batch):
|
|
grad.append(
|
|
determinant_complex_numeric_grad_single_batch(
|
|
flatten_x[b], n
|
|
)
|
|
)
|
|
else:
|
|
flatten_grad = out_grad.reshape([-1])
|
|
for b in range(n_batch):
|
|
grad.append(
|
|
determinant_complex_numeric_grad_single_batch(
|
|
flatten_x[b], n, det_out_grad=flatten_grad[b]
|
|
)
|
|
)
|
|
return np.array(grad).reshape(shape)
|
|
|
|
def test_api_static(self):
|
|
paddle.enable_static()
|
|
with paddle.static.program_guard(paddle.static.Program()):
|
|
x = paddle.static.data('X', self.shape, dtype=self.dtype)
|
|
x.stop_gradient = False
|
|
out_value = paddle.linalg.det(x)
|
|
x_grad = paddle.static.gradients([out_value], x)
|
|
exe = paddle.static.Executor(self.place)
|
|
(out_np, x_grad_np) = exe.run(
|
|
feed={'X': self.x}, fetch_list=[out_value, x_grad]
|
|
)
|
|
out_ref = np.linalg.det(self.x)
|
|
|
|
np.testing.assert_allclose(out_np, out_ref, rtol=0.001)
|
|
self.assertEqual(out_np.shape, out_ref.shape)
|
|
self.assertEqual(tuple(out_value.shape), out_ref.shape)
|
|
np.testing.assert_allclose(x_grad_np, self.x_grad_ref_st, rtol=0.001)
|
|
|
|
def test_api_dygraph(self):
|
|
paddle.disable_static(self.place)
|
|
x_tensor = paddle.to_tensor(self.x)
|
|
x_tensor.stop_gradient = False
|
|
out = paddle.linalg.det(x_tensor)
|
|
out.backward(paddle.to_tensor(self.out_grad))
|
|
out_ref = np.linalg.det(self.x)
|
|
np.testing.assert_allclose(out.numpy(), out_ref, rtol=0.001)
|
|
np.testing.assert_allclose(
|
|
x_tensor.grad.numpy(), self.x_grad_ref_dy, rtol=0.001
|
|
)
|
|
paddle.enable_static()
|
|
|
|
|
|
class TestDeterminantAPIComplex2(TestDeterminantAPIComplex):
|
|
def setUp(self):
|
|
np.random.seed(0)
|
|
self.dtype = np.complex128
|
|
self.shape = [3, 3, 5, 5]
|
|
self.x = np.vectorize(complex)(
|
|
np.random.random(self.shape), np.random.random(self.shape)
|
|
).astype(self.dtype)
|
|
self.place = paddle.CPUPlace()
|
|
self.out_grad = (
|
|
np.array([0.5 + 1.2j] * 3 * 3).reshape(3, 3).astype(self.dtype)
|
|
)
|
|
self.x_grad_ref_dy = self.get_numeric_grad(
|
|
self.x, self.shape, self.out_grad
|
|
)
|
|
self.x_grad_ref_st = self.get_numeric_grad(self.x, self.shape)
|
|
|
|
|
|
class TestSlogDeterminantOp(OpTest):
|
|
def setUp(self):
|
|
self.op_type = "slogdeterminant"
|
|
self.python_api = paddle.linalg.slogdet
|
|
self.init_data()
|
|
self.outputs = {'Out': self.target}
|
|
|
|
def test_check_output(self):
|
|
self.check_output(check_pir=True, check_symbol_infer=False)
|
|
|
|
def test_check_grad(self):
|
|
# the slog det's grad value is always huge
|
|
self.check_grad(
|
|
['Input'], ['Out'], max_relative_error=0.1, check_pir=True
|
|
)
|
|
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(4, 5, 5).astype('float64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.array(np.linalg.slogdet(self.case))
|
|
|
|
|
|
class TestSlogDeterminantOpCase1(TestSlogDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(2, 2, 5, 5).astype(np.float32)
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.array(np.linalg.slogdet(self.case))
|
|
|
|
|
|
class TestSlogDeterminantOp_ZeroSize(TestSlogDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(0, 5, 5).astype('float64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.array(np.linalg.slogdet(self.case))
|
|
|
|
|
|
class TestSlogDeterminantOp_ZeroSize2(TestSlogDeterminantOp):
|
|
def init_data(self):
|
|
np.random.seed(0)
|
|
self.case = np.random.rand(0, 0, 0).astype('float64')
|
|
self.inputs = {'Input': self.case}
|
|
self.target = np.array(np.linalg.slogdet(self.case))
|
|
|
|
|
|
class TestSlogDeterminantAPI(unittest.TestCase):
|
|
def setUp(self):
|
|
np.random.seed(0)
|
|
self.shape = [3, 3, 5, 5]
|
|
self.x = np.random.random(self.shape).astype(np.float32)
|
|
self.place = paddle.CPUPlace()
|
|
|
|
def test_api_static(self):
|
|
paddle.enable_static()
|
|
with paddle.static.program_guard(paddle.static.Program()):
|
|
x = paddle.static.data('X', self.shape)
|
|
out = paddle.linalg.slogdet(x)
|
|
exe = paddle.static.Executor(self.place)
|
|
res = exe.run(feed={'X': self.x}, fetch_list=[out])
|
|
out_ref = np.array(np.linalg.slogdet(self.x))
|
|
for out in res:
|
|
np.testing.assert_allclose(out, out_ref, rtol=0.001)
|
|
|
|
def test_api_dygraph(self):
|
|
paddle.disable_static(self.place)
|
|
x_tensor = paddle.to_tensor(self.x)
|
|
out = paddle.linalg.slogdet(x_tensor)
|
|
out_ref = np.array(np.linalg.slogdet(self.x))
|
|
np.testing.assert_allclose(out.numpy(), out_ref, rtol=0.001)
|
|
paddle.enable_static()
|
|
|
|
|
|
def slogdeterminant_complex_numeric_grad_single_batch(
|
|
x, n, delta=0.005, logabsdet_out_grad=np.array(1 + 0j)
|
|
):
|
|
# an naive implementation of numeric_grad with single batch input x
|
|
# the output of logabsdet is always real, so logabsdet_out_grad
|
|
# should be a+0j, where a is an arbitrary real number
|
|
dx = []
|
|
for i in range(n):
|
|
for j in range(n):
|
|
xp = x.copy()
|
|
xn = x.copy()
|
|
xpj = x.copy()
|
|
xnj = x.copy()
|
|
xp[i, j] += delta
|
|
xn[i, j] -= delta
|
|
xpj[i, j] += delta * 1j
|
|
xnj[i, j] -= delta * 1j
|
|
_, yp = np.linalg.slogdet(xp)
|
|
_, yn = np.linalg.slogdet(xn)
|
|
_, ypj = np.linalg.slogdet(xpj)
|
|
_, ynj = np.linalg.slogdet(xnj)
|
|
df_over_dr = (yp - yn) / delta / 2
|
|
df_over_di = (ypj - ynj) / delta / 2
|
|
dl_over_du, dl_over_dv = (
|
|
logabsdet_out_grad.real,
|
|
logabsdet_out_grad.imag,
|
|
)
|
|
du_over_dr, dv_over_dr = df_over_dr.real, df_over_dr.imag
|
|
du_over_di, dv_over_di = df_over_di.real, df_over_di.imag
|
|
dl_over_dr = np.sum(
|
|
dl_over_du * du_over_dr + dl_over_dv * dv_over_dr
|
|
)
|
|
dl_over_di = np.sum(
|
|
dl_over_du * du_over_di + dl_over_dv * dv_over_di
|
|
)
|
|
dx.append(dl_over_dr + 1j * dl_over_di)
|
|
return np.array(dx).reshape([n, n])
|
|
|
|
|
|
class TestSlogDeterminantAPIComplex(unittest.TestCase):
|
|
def setUp(self):
|
|
np.random.seed(0)
|
|
self.shape = [3, 3, 5, 5]
|
|
self.dtype = np.complex64
|
|
self.x = np.vectorize(complex)(
|
|
np.random.random(self.shape), np.random.random(self.shape)
|
|
).astype(self.dtype)
|
|
self.places = get_places()
|
|
self.out_grad = (
|
|
np.array([1 + 0j, 1 + 0j] * 3 * 3)
|
|
.reshape(2, 3, 3)
|
|
.astype(self.dtype)
|
|
)
|
|
self.x_grad_ref_dy = self.get_numeric_grad(
|
|
self.x, self.shape, self.out_grad
|
|
)
|
|
self.x_grad_ref_st = self.get_numeric_grad(self.x, self.shape)
|
|
|
|
def get_numeric_grad(self, x, shape, out_grad=None):
|
|
n = shape[-1]
|
|
flatten_x = x.reshape([-1, n, n])
|
|
n_batch = flatten_x.shape[0]
|
|
grad = []
|
|
if out_grad is None:
|
|
for b in range(n_batch):
|
|
grad.append(
|
|
slogdeterminant_complex_numeric_grad_single_batch(
|
|
flatten_x[b], n
|
|
)
|
|
)
|
|
else:
|
|
flatten_grad = out_grad.reshape([-1, 2])
|
|
for b in range(n_batch):
|
|
grad.append(
|
|
slogdeterminant_complex_numeric_grad_single_batch(
|
|
flatten_x[b], n, logabsdet_out_grad=flatten_grad[b][1]
|
|
)
|
|
)
|
|
return np.array(grad).reshape(shape)
|
|
|
|
def test_api_static(self):
|
|
for place in self.places:
|
|
paddle.enable_static()
|
|
with paddle.static.program_guard(paddle.static.Program()):
|
|
x = paddle.static.data('X', self.shape, self.dtype)
|
|
x.stop_gradient = False
|
|
out = paddle.linalg.slogdet(x)
|
|
x_grad = paddle.static.gradients(out, x)
|
|
exe = paddle.static.Executor(place)
|
|
res = exe.run(feed={'X': self.x}, fetch_list=[out, x_grad])
|
|
out_ref = np.array(np.linalg.slogdet(self.x))
|
|
np.testing.assert_allclose(res[0], out_ref, rtol=0.001)
|
|
np.testing.assert_allclose(res[1], self.x_grad_ref_st, rtol=0.001)
|
|
|
|
def test_api_dygraph(self):
|
|
for place in self.places:
|
|
paddle.disable_static(place)
|
|
x_tensor = paddle.to_tensor(self.x)
|
|
x_tensor.stop_gradient = False
|
|
out = paddle.linalg.slogdet(x_tensor)
|
|
out.backward(paddle.to_tensor(self.out_grad))
|
|
out_ref = np.array(np.linalg.slogdet(self.x))
|
|
np.testing.assert_allclose(out.numpy(), out_ref, rtol=0.001)
|
|
np.testing.assert_allclose(
|
|
x_tensor.grad.numpy(), self.x_grad_ref_dy, rtol=0.001
|
|
)
|
|
paddle.enable_static()
|
|
|
|
|
|
class TestSlogDeterminantAPIComplex2(TestSlogDeterminantAPIComplex):
|
|
def setUp(self):
|
|
np.random.seed(0)
|
|
self.shape = [6, 5, 5]
|
|
self.dtype = np.complex128
|
|
self.x = np.vectorize(complex)(
|
|
np.random.random(self.shape), np.random.random(self.shape)
|
|
).astype(self.dtype)
|
|
self.places = get_places()
|
|
self.out_grad = np.array([3 + 0j, 3 + 0j] * 6).reshape(2, 6)
|
|
self.x_grad_ref_dy = self.get_numeric_grad(
|
|
self.x, self.shape, self.out_grad
|
|
)
|
|
self.x_grad_ref_st = self.get_numeric_grad(self.x, self.shape)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|