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
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
|
from __future__ import annotations
import string
import numpy as np
import pandas as pd
import pytest
from anndata._core.xarray import Dataset2D
from anndata.compat import XDataArray, XDataset, XVariable
from anndata.tests.helpers import gen_typed_df
pytest.importorskip("xarray")
@pytest.fixture
def df():
return gen_typed_df(10)
@pytest.fixture
def dataset2d(df):
return Dataset2D(XDataset.from_dataframe(df))
def test_shape(df, dataset2d):
assert dataset2d.shape == df.shape
def test_columns(df, dataset2d):
assert np.all(dataset2d.columns.sort_values() == df.columns.sort_values())
def test_to_memory(df, dataset2d):
memory_df = dataset2d.to_memory()
assert np.all(df == memory_df)
assert np.all(df.index == memory_df.index)
assert np.all(df.columns.sort_values() == memory_df.columns.sort_values())
def test_getitem(df, dataset2d):
col = df.columns[0]
assert np.all(dataset2d[col] == df[col])
empty_dset = dataset2d[[]]
assert empty_dset.shape == (df.shape[0], 0)
assert np.all(empty_dset.index == dataset2d.index)
def test_backed_property(dataset2d):
assert not dataset2d.is_backed
dataset2d.is_backed = True
assert dataset2d.is_backed
dataset2d.is_backed = False
assert not dataset2d.is_backed
def test_index_dim(dataset2d):
assert dataset2d.index_dim == "index"
assert dataset2d.true_index_dim == dataset2d.index_dim
col = next(iter(dataset2d.keys()))
dataset2d.true_index_dim = col
assert dataset2d.index_dim == "index"
assert dataset2d.true_index_dim == col
with pytest.raises(ValueError, match=r"Unknown variable `test`\."):
dataset2d.true_index_dim = "test"
dataset2d.true_index_dim = None
assert dataset2d.true_index_dim == dataset2d.index_dim
def test_index(dataset2d):
alphabet = np.asarray(
list(string.ascii_letters + string.digits + string.punctuation)
)
new_idx = pd.Index(
[
"".join(np.random.choice(alphabet, size=10))
for _ in range(dataset2d.shape[0])
],
name="test_index",
)
col = next(iter(dataset2d.keys()))
dataset2d.true_index_dim = col
dataset2d.index = new_idx
assert np.all(dataset2d.index == new_idx)
assert dataset2d.true_index_dim == dataset2d.index_dim == new_idx.name
assert list(dataset2d.ds.coords.keys()) == [new_idx.name]
@pytest.fixture
def dataset_2d_one_column():
return Dataset2D(
XDataset(
{"foo": ("obs_names", pd.array(["a", "b", "c"], dtype="category"))},
coords={"obs_names": [1, 2, 3]},
)
)
def test_dataset_2d_set_dataarray(dataset_2d_one_column):
da = XDataArray(
np.arange(3), coords={"obs_names": [1, 2, 3]}, dims=("obs_names"), name="bar"
)
dataset_2d_one_column["bar"] = da
assert dataset_2d_one_column["bar"].dims == ("obs_names",)
assert dataset_2d_one_column["bar"].equals(da)
def test_dataset_2d_set_dataset(dataset_2d_one_column):
ds = XDataset(
data_vars={
"foo": ("obs_names", np.arange(3)),
"bar": ("obs_names", np.arange(3) + 3),
},
coords={"obs_names": [1, 2, 3]},
)
key = ["foo", "bar"]
dataset_2d_one_column[key] = ds
assert tuple(dataset_2d_one_column[key].ds.sizes.keys()) == ("obs_names",)
assert dataset_2d_one_column[key].equals(ds)
@pytest.mark.parametrize(
"setter",
[
pd.array(["e", "f", "g"], dtype="category"),
("obs_names", pd.array(["e", "f", "g"], dtype="category")),
],
ids=["array", "tuple_with_array"],
)
def test_dataset_2d_set_extension_array(dataset_2d_one_column, setter):
dataset_2d_one_column["bar"] = setter
assert dataset_2d_one_column["bar"].dims == ("obs_names",)
assert (
dataset_2d_one_column["bar"].data is setter[1]
if isinstance(setter, tuple)
else setter
)
@pytest.mark.parametrize(
("da", "pattern"),
[
pytest.param(
XDataset(
data_vars={"bar": ("obs_names", np.arange(3))},
coords={"foo": ("obs_names", np.arange(3))},
),
r"Dataset should have coordinate obs_names",
id="coord_name_dataset",
),
pytest.param(
XDataArray(
np.arange(3),
coords={"foo": ("obs_names", np.arange(3))},
dims="obs_names",
name="bar",
),
r"DataArray should have coordinate obs_names",
id="coord_name",
),
pytest.param(
XDataArray(
np.arange(3),
coords={"obs_names": np.arange(3)},
dims=("obs_names",),
name="not_bar",
),
r"DataArray should have name bar, found not_bar",
id="dataarray_name",
),
pytest.param(
XDataset(
data_vars={
"foo": (["obs_names", "not_obs_names"], np.arange(9).reshape(3, 3))
},
coords={"obs_names": np.arange(3), "not_obs_names": np.arange(3)},
),
r"Dataset should have only one dimension",
id="multiple_dims_dataset",
),
pytest.param(
XDataArray(
np.arange(9).reshape(3, 3),
coords={"obs_names": np.arange(3), "not_obs_names": np.arange(3)},
dims=("obs_names", "not_obs_names"),
),
r"DataArray should have only one dimension",
id="multiple_dims_dataarray",
),
pytest.param(
XVariable(
data=np.arange(9).reshape(3, 3),
dims=("obs_names", "not_obs_names"),
),
r"Variable should have only one dimension",
id="multiple_dims_variable",
),
pytest.param(
XDataset(
data_vars={"foo": ("other", np.arange(3))},
coords={"obs_names": ("other", np.arange(3))},
),
r"Dataset should have dimension obs_names",
id="name_conflict_dataset",
),
pytest.param(
XVariable(
data=np.arange(3),
dims="not_obs_names",
),
r"Variable should have dimension obs_names, found not_obs_names",
id="name_conflict_variable",
),
pytest.param(
XDataArray(
np.arange(3),
coords=[np.arange(3)],
dims="not_obs_names",
),
r"DataArray should have dimension obs_names, found not_obs_names",
id="name_conflict_dataarray",
),
pytest.param(
("not_obs_names", [1, 2, 3]),
r"Setting value tuple should have first entry",
id="tuple_bad_dim",
),
pytest.param(
(("not_obs_names",), [1, 2, 3]),
r"Dimension tuple should have only",
id="nested_tuple_bad_dim",
),
pytest.param(
(("obs_names", "bar"), [1, 2, 3]),
r"Dimension tuple is too long",
id="nested_tuple_too_long",
),
],
)
def test_dataset_2d_set_with_bad_obj(da, pattern, dataset_2d_one_column):
with pytest.raises(ValueError, match=pattern):
dataset_2d_one_column["bar"] = da
@pytest.mark.parametrize(
"data", [np.arange(3), XDataArray(np.arange(3), dims="obs_names", name="obs_names")]
)
def test_dataset_2d_set_index(data, dataset_2d_one_column):
with pytest.raises(
KeyError,
match="Cannot set obs_names as a variable",
):
dataset_2d_one_column["obs_names"] = data
@pytest.mark.parametrize(
("ds", "pattern", "error"),
[
pytest.param(
XDataset(
{"foo": ("obs_names", pd.array(["a", "b", "c"], dtype="category"))},
coords={"obs_names": ("not_obs_names", [1, 2, 3])},
),
r"Dataset should have exactly one dimension",
ValueError,
id="more_than_one_dimension",
),
pytest.param(
XDataset(
{"foo": ("obs_names", pd.array(["a", "b", "c"], dtype="category"))},
coords={
"obs_names": ("obs_names", [1, 2, 3]),
"not_obs_names": ("obs_names", [1, 2, 3]),
},
),
r"Dataset should have exactly one coordinate",
ValueError,
id="more_than_one_coord",
),
pytest.param(
XDataset(
{"foo": ("not_obs_names", pd.array(["a", "b", "c"], dtype="category"))},
coords={
"obs_names": ("not_obs_names", [1, 2, 3]),
},
),
r"does not match coordinate",
ValueError,
id="coord_dim_mismatch",
),
pytest.param(
XDataset(
{"foo": (("obs", "obs1"), np.arange(9).reshape(3, 3))},
coords={
"obs_names": (("obs", "obs1"), np.arange(9).reshape(3, 3)),
},
),
r"Dataset should have exactly one",
ValueError,
id="multi_dim_coord",
),
pytest.param(
dict(foo="bar"),
r"Expected an xarray Dataset",
TypeError,
id="non_ds_init",
),
],
)
def test_init_errors(ds, pattern, error):
with pytest.raises(error, match=pattern):
Dataset2D(ds)
|
