# Copyright (c) 2024 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 get_device_place import paddle np.random.seed(10) paddle.seed(10) def log_normal_mean(mean, std): return np.exp(mean + np.power(std, 2) / 2.0) def log_normal_var(mean, std): var = np.power(std, 2) return (np.exp(var) - 1.0) * np.exp(2 * mean + var) class TestLogNormalAPI(unittest.TestCase): def setUp(self): self.mean = 0.0 self.std = 0.5 self.shape = None self.mean_duplicate = None self.std_duplicate = None self.duplicates = 1000 self.set_attrs() self.dtype = self.get_dtype() self.place = get_device_place() def set_attrs(self): self.shape = [self.duplicates] def get_shape(self): if isinstance(self.mean, np.ndarray): shape = self.mean_duplicate.shape elif isinstance(self.std, np.ndarray): shape = self.std_duplicate.shape else: shape = self.shape return list(shape) def get_dtype(self): if isinstance(self.mean, np.ndarray): return self.mean.dtype elif isinstance(self.std, np.ndarray): return self.std.dtype else: return 'float32' def static_api(self): shape = self.get_shape() main_program = paddle.static.Program() if isinstance(self.mean, np.ndarray) and isinstance( self.std, np.ndarray ): with paddle.static.program_guard(main_program): mean = paddle.static.data( 'Mean', self.mean_duplicate.shape, self.mean_duplicate.dtype ) std = paddle.static.data( 'Std', self.std_duplicate.shape, self.std_duplicate.dtype ) out = paddle.log_normal(mean, std, self.shape) exe = paddle.static.Executor(self.place) ret = exe.run( feed={ 'Mean': self.mean_duplicate, 'Std': self.std_duplicate.reshape(shape), }, fetch_list=[out], ) return ret[0] elif isinstance(self.mean, np.ndarray): with paddle.static.program_guard(main_program): mean = paddle.static.data( 'Mean', self.mean_duplicate.shape, self.mean_duplicate.dtype ) out = paddle.log_normal(mean, self.std, self.shape) exe = paddle.static.Executor(self.place) ret = exe.run( feed={'Mean': self.mean_duplicate}, fetch_list=[out] ) return ret[0] elif isinstance(self.std, np.ndarray): with paddle.static.program_guard(main_program): std = paddle.static.data( 'Std', self.std_duplicate.shape, self.std_duplicate.dtype ) out = paddle.log_normal(self.mean, std, self.shape) exe = paddle.static.Executor(self.place) ret = exe.run( feed={'Std': self.std_duplicate}, fetch_list=[out] ) return ret[0] else: with paddle.static.program_guard(main_program): out = paddle.log_normal(self.mean, self.std, self.shape) exe = paddle.static.Executor(self.place) ret = exe.run(fetch_list=[out]) return ret[0] def dygraph_api(self): paddle.disable_static(self.place) shape = self.get_shape() mean = ( paddle.to_tensor(self.mean_duplicate) if isinstance(self.mean, np.ndarray) else self.mean ) std = ( paddle.to_tensor(self.std_duplicate) if isinstance(self.std, np.ndarray) else self.std ) out = paddle.log_normal(mean, std, self.shape) ret = out.numpy() paddle.enable_static() return ret def test_api(self): paddle.enable_static() ret_static = self.static_api() ret_dygraph = self.dygraph_api() for ret in [ret_static, ret_dygraph]: shape_ref = self.get_shape() self.assertEqual(shape_ref, list(ret.shape)) mean = np.mean(ret, axis=0, keepdims=True) var = np.var(ret, axis=0, keepdims=True) mean_ref = log_normal_mean(self.mean, self.std) var_ref = log_normal_var(self.mean, self.std) np.testing.assert_allclose(mean_ref, mean, rtol=0.2, atol=0.2) np.testing.assert_allclose(var_ref, var, rtol=0.2, atol=0.2) class TestLogNormalAPI_mean_is_tensor(TestLogNormalAPI): def set_attrs(self): self.mean = np.random.uniform(-0.5, -0.1, [1, 2]).astype('float64') self.mean_duplicate = np.broadcast_to(self.mean, [self.duplicates, 2]) self.std = 0.5 class TestLogNormalAPI_std_is_tensor(TestLogNormalAPI): def set_attrs(self): self.std = np.random.uniform(0.1, 0.5, [1, 2]).astype('float64') self.std_duplicate = np.broadcast_to(self.std, [self.duplicates, 2]) class TestLogNormalAPI_mean_std_are_tensor(TestLogNormalAPI): def set_attrs(self): self.mean = np.random.uniform(0.1, 0.5, [1, 2]).astype('float64') self.mean_duplicate = np.broadcast_to(self.mean, [self.duplicates, 2]) self.std = np.random.uniform(0.1, 0.5, [1, 2]).astype('float64') self.std_duplicate = np.broadcast_to(self.std, [self.duplicates, 2]) class TestLogNormalAlias(unittest.TestCase): def test_alias(self): paddle.disable_static() shape = [1, 2, 3] out1 = paddle.log_normal(shape=shape) out2 = paddle.tensor.log_normal(shape=shape) out3 = paddle.tensor.random.log_normal(shape=shape) paddle.enable_static() class TestLogNormalErrors(unittest.TestCase): def test_errors(self): main_program = paddle.static.Program() with paddle.static.program_guard(main_program): mean = [1, 2, 3] self.assertRaises(TypeError, paddle.log_normal, mean) std = [1, 2, 3] self.assertRaises(TypeError, paddle.log_normal, std=std) mean = paddle.static.data('Mean', [100], 'int32') self.assertRaises(TypeError, paddle.log_normal, mean) std = paddle.static.data('Std', [100], 'int32') self.assertRaises(TypeError, paddle.log_normal, mean=1.0, std=std) self.assertRaises(TypeError, paddle.log_normal, shape=1) self.assertRaises(TypeError, paddle.log_normal, shape=[1.0]) shape = paddle.static.data('Shape', [100], 'float32') self.assertRaises(TypeError, paddle.log_normal, shape=shape) if __name__ == "__main__": unittest.main()