wehub-resource-sync
148 lines
5.2 KiB
Python
148 lines
5.2 KiB
Python
# Copyright (c) 2025 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 paddle
|
|
|
|
|
|
class TestEmptyTensorIndexing(unittest.TestCase):
|
|
"""Test empty tensor indexing operations."""
|
|
|
|
def test_consecutive_empty_tensor_indexing(self):
|
|
"""Test consecutive indexing on empty tensors.
|
|
|
|
This test case reproduces the bug reported in issue #76194.
|
|
Before the fix, the second consecutive empty tensor indexing
|
|
operation would fail with ValueError.
|
|
"""
|
|
cum_nodes = paddle.zeros([2])
|
|
batch = paddle.empty([0], dtype=paddle.int64)
|
|
edge_index = paddle.empty([2, 0], dtype=paddle.int64)
|
|
|
|
# Both operations should succeed
|
|
result1 = cum_nodes[batch][edge_index[0]]
|
|
result2 = cum_nodes[batch][edge_index[1]]
|
|
|
|
# Verify results
|
|
self.assertEqual(result1.shape, [0])
|
|
self.assertEqual(result2.shape, [0])
|
|
self.assertEqual(result1.dtype, cum_nodes.dtype)
|
|
self.assertEqual(result2.dtype, cum_nodes.dtype)
|
|
|
|
def test_multiple_consecutive_empty_indexing(self):
|
|
"""Test multiple consecutive empty tensor indexing operations."""
|
|
tensor = paddle.zeros([10])
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
# Perform multiple consecutive indexing operations
|
|
for i in range(10):
|
|
result = tensor[empty_idx]
|
|
self.assertEqual(result.shape, [0])
|
|
self.assertEqual(result.dtype, tensor.dtype)
|
|
|
|
def test_empty_2d_tensor_indexing(self):
|
|
"""Test 2D empty tensor indexing."""
|
|
tensor = paddle.zeros([3, 4])
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
result = tensor[empty_idx]
|
|
self.assertEqual(result.shape, [0, 4])
|
|
self.assertEqual(result.dtype, tensor.dtype)
|
|
|
|
def test_empty_3d_tensor_indexing(self):
|
|
"""Test 3D empty tensor indexing."""
|
|
tensor = paddle.zeros([5, 5, 5])
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
result1 = tensor[empty_idx]
|
|
self.assertEqual(result1.shape, [0, 5, 5])
|
|
|
|
# Test chained empty indexing
|
|
result2 = tensor[empty_idx][empty_idx]
|
|
self.assertEqual(result2.shape[0], 0)
|
|
|
|
def test_mixed_empty_nonempty_indexing(self):
|
|
"""Test alternating empty and non-empty tensor indexing."""
|
|
tensor = paddle.zeros([10])
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
normal_idx = paddle.to_tensor([0, 1, 2], dtype=paddle.int64)
|
|
|
|
# Alternate between empty and non-empty indexing
|
|
for i in range(5):
|
|
if i % 2 == 0:
|
|
result = tensor[empty_idx]
|
|
self.assertEqual(result.shape, [0])
|
|
else:
|
|
result = tensor[normal_idx]
|
|
self.assertEqual(result.shape, [3])
|
|
|
|
def test_empty_tensor_with_different_dtypes(self):
|
|
"""Test empty tensor indexing with different data types."""
|
|
dtypes = [
|
|
paddle.float32,
|
|
paddle.float64,
|
|
paddle.int32,
|
|
paddle.int64,
|
|
]
|
|
|
|
for dtype in dtypes:
|
|
tensor = paddle.zeros([5], dtype=dtype)
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
result = tensor[empty_idx]
|
|
self.assertEqual(result.shape, [0])
|
|
self.assertEqual(result.dtype, dtype)
|
|
|
|
def test_chained_triple_empty_indexing(self):
|
|
"""Test three consecutive empty indexing operations."""
|
|
tensor = paddle.zeros([10, 10, 10])
|
|
idx1 = paddle.empty([0], dtype=paddle.int64)
|
|
idx2 = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
try:
|
|
result = tensor[idx1]
|
|
result = result[idx2]
|
|
self.assertEqual(result.shape[0], 0)
|
|
except Exception as e:
|
|
self.fail(f"Chained empty indexing failed with: {e}")
|
|
|
|
|
|
class TestEmptyTensorIndexingEdgeCases(unittest.TestCase):
|
|
"""Test edge cases for empty tensor indexing."""
|
|
|
|
def test_empty_index_on_empty_tensor(self):
|
|
"""Test indexing an empty tensor with an empty index."""
|
|
empty_tensor = paddle.empty([0], dtype=paddle.float32)
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
result = empty_tensor[empty_idx]
|
|
self.assertEqual(result.shape, [0])
|
|
|
|
def test_empty_tensor_holder_size(self):
|
|
"""Test that empty tensor holder is correctly initialized."""
|
|
tensor = paddle.zeros([5])
|
|
empty_idx = paddle.empty([0], dtype=paddle.int64)
|
|
|
|
# Multiple indexing operations should not corrupt holder
|
|
for _ in range(20):
|
|
result = tensor[empty_idx]
|
|
# Verify the tensor is still valid
|
|
self.assertEqual(result.numel(), 0)
|
|
self.assertTrue(result.is_tensor())
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|