forked from huggingface/datasets
-
Notifications
You must be signed in to change notification settings - Fork 0
/
test_py_utils.py
147 lines (128 loc) · 5.86 KB
/
test_py_utils.py
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
from unittest import TestCase
import numpy as np
import pytest
from datasets.utils.py_utils import (
NestedDataStructure,
flatten_nest_dict,
map_nested,
temporary_assignment,
zip_dict,
zip_nested,
)
def np_sum(x): # picklable for multiprocessing
return x.sum()
def add_one(i): # picklable for multiprocessing
return i + 1
class PyUtilsTest(TestCase):
def test_flatten_nest_dict(self):
d1 = {}
d2 = {"a": 1, "b": 2}
d3 = {"a": {"1": 1, "2": 2}, "b": 3}
expected_flatten_d1 = {}
expected_flatten_d2 = {"a": 1, "b": 2}
expected_flatten_d3 = {"a/1": 1, "a/2": 2, "b": 3}
self.assertDictEqual(flatten_nest_dict(d1), expected_flatten_d1)
self.assertDictEqual(flatten_nest_dict(d2), expected_flatten_d2)
self.assertDictEqual(flatten_nest_dict(d3), expected_flatten_d3)
def test_map_nested(self):
s1 = {}
s2 = []
s3 = 1
s4 = [1, 2]
s5 = {"a": 1, "b": 2}
s6 = {"a": [1, 2], "b": [3, 4]}
s7 = {"a": {"1": 1}, "b": 2}
s8 = {"a": 1, "b": 2, "c": 3, "d": 4}
expected_map_nested_s1 = {}
expected_map_nested_s2 = []
expected_map_nested_s3 = 2
expected_map_nested_s4 = [2, 3]
expected_map_nested_s5 = {"a": 2, "b": 3}
expected_map_nested_s6 = {"a": [2, 3], "b": [4, 5]}
expected_map_nested_s7 = {"a": {"1": 2}, "b": 3}
expected_map_nested_s8 = {"a": 2, "b": 3, "c": 4, "d": 5}
self.assertEqual(map_nested(add_one, s1), expected_map_nested_s1)
self.assertEqual(map_nested(add_one, s2), expected_map_nested_s2)
self.assertEqual(map_nested(add_one, s3), expected_map_nested_s3)
self.assertEqual(map_nested(add_one, s4), expected_map_nested_s4)
self.assertEqual(map_nested(add_one, s5), expected_map_nested_s5)
self.assertEqual(map_nested(add_one, s6), expected_map_nested_s6)
self.assertEqual(map_nested(add_one, s7), expected_map_nested_s7)
self.assertEqual(map_nested(add_one, s8), expected_map_nested_s8)
num_proc = 2
self.assertEqual(map_nested(add_one, s1, num_proc=num_proc), expected_map_nested_s1)
self.assertEqual(map_nested(add_one, s2, num_proc=num_proc), expected_map_nested_s2)
self.assertEqual(map_nested(add_one, s3, num_proc=num_proc), expected_map_nested_s3)
self.assertEqual(map_nested(add_one, s4, num_proc=num_proc), expected_map_nested_s4)
self.assertEqual(map_nested(add_one, s5, num_proc=num_proc), expected_map_nested_s5)
self.assertEqual(map_nested(add_one, s6, num_proc=num_proc), expected_map_nested_s6)
self.assertEqual(map_nested(add_one, s7, num_proc=num_proc), expected_map_nested_s7)
self.assertEqual(map_nested(add_one, s8, num_proc=num_proc), expected_map_nested_s8)
sn1 = {"a": np.eye(2), "b": np.zeros(3), "c": np.ones(2)}
expected_map_nested_sn1_sum = {"a": 2, "b": 0, "c": 2}
expected_map_nested_sn1_int = {
"a": np.eye(2).astype(int),
"b": np.zeros(3).astype(int),
"c": np.ones(2).astype(int),
}
self.assertEqual(map_nested(np_sum, sn1, map_numpy=False), expected_map_nested_sn1_sum)
self.assertEqual(
{k: v.tolist() for k, v in map_nested(int, sn1, map_numpy=True).items()},
{k: v.tolist() for k, v in expected_map_nested_sn1_int.items()},
)
self.assertEqual(map_nested(np_sum, sn1, map_numpy=False, num_proc=num_proc), expected_map_nested_sn1_sum)
self.assertEqual(
{k: v.tolist() for k, v in map_nested(int, sn1, map_numpy=True, num_proc=num_proc).items()},
{k: v.tolist() for k, v in expected_map_nested_sn1_int.items()},
)
with self.assertRaises(AttributeError): # can't pickle a local lambda
map_nested(lambda x: x + 1, sn1, num_proc=num_proc)
def test_zip_dict(self):
d1 = {"a": 1, "b": 2}
d2 = {"a": 3, "b": 4}
d3 = {"a": 5, "b": 6}
expected_zip_dict_result = sorted([("a", (1, 3, 5)), ("b", (2, 4, 6))])
self.assertEqual(sorted(list(zip_dict(d1, d2, d3))), expected_zip_dict_result)
def test_zip_nested(self):
d1 = {"a": {"1": 1}, "b": 2}
d2 = {"a": {"1": 3}, "b": 4}
expected_zip_nested_result = {"a": {"1": (1, 3)}, "b": (2, 4)}
self.assertDictEqual(zip_nested(d1, d2), expected_zip_nested_result)
def test_temporary_assignment(self):
class Foo:
my_attr = "bar"
foo = Foo()
self.assertEqual(foo.my_attr, "bar")
with temporary_assignment(foo, "my_attr", "BAR"):
self.assertEqual(foo.my_attr, "BAR")
self.assertEqual(foo.my_attr, "bar")
@pytest.mark.parametrize("input_data", [{}])
def test_nested_data_structure_data(input_data):
output_data = NestedDataStructure(input_data).data
assert output_data == input_data
@pytest.mark.parametrize(
"data, expected_output",
[
({}, []),
([], []),
("foo", ["foo"]),
(["foo", "bar"], ["foo", "bar"]),
([["foo", "bar"]], ["foo", "bar"]),
([[["foo"], ["bar"]]], ["foo", "bar"]),
([[["foo"], "bar"]], ["foo", "bar"]),
({"a": 1, "b": 2}, [1, 2]),
({"a": [1, 2], "b": [3, 4]}, [1, 2, 3, 4]),
({"a": [[1, 2]], "b": [[3, 4]]}, [1, 2, 3, 4]),
({"a": [[1, 2]], "b": [3, 4]}, [1, 2, 3, 4]),
({"a": [[[1], [2]]], "b": [[[3], [4]]]}, [1, 2, 3, 4]),
({"a": [[[1], [2]]], "b": [[3, 4]]}, [1, 2, 3, 4]),
({"a": [[[1], [2]]], "b": [3, 4]}, [1, 2, 3, 4]),
({"a": [[[1], [2]]], "b": [3, [4]]}, [1, 2, 3, 4]),
({"a": {"1": 1}, "b": 2}, [1, 2]),
({"a": {"1": [1]}, "b": 2}, [1, 2]),
({"a": {"1": [1]}, "b": [2]}, [1, 2]),
],
)
def test_flatten(data, expected_output):
output = NestedDataStructure(data).flatten()
assert output == expected_output