1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
|
# ----------------------------------------------------------------------------
# - Open3D: www.open3d.org -
# ----------------------------------------------------------------------------
# Copyright (c) 2018-2024 www.open3d.org
# SPDX-License-Identifier: MIT
# ----------------------------------------------------------------------------
import open3d as o3d
import numpy as np
import pytest
import mltest
# skip all tests if the ml ops were not built
pytestmark = mltest.default_marks
# the supported dtypes for the attributes
value_dtypes = pytest.mark.parametrize(
'dtype',
[np.uint8, np.int8, np.int16, np.int32, np.int64, np.float32, np.float64])
attributes = pytest.mark.parametrize('attributes',
['scalar', 'none', 'multidim'])
@value_dtypes
@attributes
@mltest.parametrize.ml
def test_invert_neighbors_list(dtype, attributes, ml):
# yapf: disable
# define connectivity for 3 query points and 3 input points
num_points = 3
edges = np.array(
[
[0, 0], [0, 1], [0, 2], # 3 neighbors
[1, 2], # 1 neighbors
[2, 1], [2, 2], # 2 neighbors
],
dtype=np.int32)
# the neighbors_index is the second column
neighbors_index = edges[:, 1]
# exclusive prefix sum of the number of neighbors
neighbors_row_splits = np.array([0, 3, 4, edges.shape[0]], dtype=np.int64)
if attributes == 'scalar':
neighbors_attributes = np.array([
10, 20, 30,
40,
50, 60,
], dtype=dtype)
elif attributes == 'none':
neighbors_attributes = np.array([], dtype=dtype)
elif attributes == 'multidim':
neighbors_attributes = np.array([
[10, 1], [20, 2], [30, 3],
[40, 4],
[50, 5], [60, 6],
], dtype=dtype)
# yapf: enable
ans = mltest.run_op(ml,
ml.device,
True,
ml.ops.invert_neighbors_list,
num_points=num_points,
inp_neighbors_index=neighbors_index,
inp_neighbors_row_splits=neighbors_row_splits,
inp_neighbors_attributes=neighbors_attributes)
expected_neighbors_row_splits = [0, 1, 3, edges.shape[0]]
np.testing.assert_equal(ans.neighbors_row_splits,
expected_neighbors_row_splits)
# checking the neighbors_index is more complicated because the order
# of the neighbors for each query point is not defined.
expected_neighbors_index = [
set([0]),
set([0, 2]),
set([0, 1, 2]),
]
for i, expected_neighbors_i in enumerate(expected_neighbors_index):
start = ans.neighbors_row_splits[i]
end = ans.neighbors_row_splits[i + 1]
neighbors_i = set(ans.neighbors_index[start:end])
assert neighbors_i == expected_neighbors_i
if neighbors_attributes.shape == (0,):
# if the input is a zero length vector then the returned attributes
# vector also must be a zero length vector
assert ans.neighbors_attributes.shape == (0,)
else:
# check if the attributes are still associated with the same edge
edge_attr_map = {
tuple(k): v for k, v in zip(edges, neighbors_attributes)
}
for i, _ in enumerate(expected_neighbors_index):
start = ans.neighbors_row_splits[i]
end = ans.neighbors_row_splits[i + 1]
# neighbors and attributes for point i
neighbors_i = ans.neighbors_index[start:end]
attributes_i = ans.neighbors_attributes[start:end]
for j, attr in zip(neighbors_i, attributes_i):
key = (j, i)
np.testing.assert_equal(attr, edge_attr_map[key])
@mltest.parametrize.ml
def test_invert_neighbors_list_shape_checking(ml):
num_points = 3
inp_neighbors_index = np.array([0, 1, 2, 2, 1, 2], dtype=np.int32)
inp_neighbors_row_splits = np.array([0, 3, 4, 6], dtype=np.int64)
inp_neighbors_attributes = np.array([10, 20, 30, 40, 50, 60],
dtype=np.float32)
# test the shape checking by passing arrays with wrong rank and/or size
with pytest.raises(Exception) as einfo:
_ = mltest.run_op(ml,
ml.cpu_device,
False,
ml.ops.invert_neighbors_list,
num_points=num_points,
inp_neighbors_index=inp_neighbors_index[1:],
inp_neighbors_row_splits=inp_neighbors_row_splits,
inp_neighbors_attributes=inp_neighbors_attributes)
assert 'invalid shape' in str(einfo.value)
with pytest.raises(Exception) as einfo:
_ = mltest.run_op(ml,
ml.cpu_device,
False,
ml.ops.invert_neighbors_list,
num_points=num_points,
inp_neighbors_index=inp_neighbors_index[:,
np.newaxis],
inp_neighbors_row_splits=inp_neighbors_row_splits,
inp_neighbors_attributes=inp_neighbors_attributes)
assert 'invalid shape' in str(einfo.value)
with pytest.raises(Exception) as einfo:
_ = mltest.run_op(
ml,
ml.cpu_device,
False,
ml.ops.invert_neighbors_list,
num_points=num_points,
inp_neighbors_index=inp_neighbors_index,
inp_neighbors_row_splits=inp_neighbors_row_splits[:, np.newaxis],
inp_neighbors_attributes=inp_neighbors_attributes)
assert 'invalid shape' in str(einfo.value)
with pytest.raises(Exception) as einfo:
_ = mltest.run_op(ml,
ml.cpu_device,
False,
ml.ops.invert_neighbors_list,
num_points=num_points,
inp_neighbors_index=inp_neighbors_index,
inp_neighbors_row_splits=inp_neighbors_row_splits,
inp_neighbors_attributes=inp_neighbors_attributes[1:])
assert 'invalid shape' in str(einfo.value)
|
