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
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
|
import numpy as np
import pytest
import xarray as xr
from xarray import DataArray
from xarray.tests import assert_allclose, assert_equal, raises_regex
from . import raise_if_dask_computes, requires_dask
@pytest.mark.parametrize("as_dataset", (True, False))
def test_weighted_non_DataArray_weights(as_dataset):
data = DataArray([1, 2])
if as_dataset:
data = data.to_dataset(name="data")
with raises_regex(ValueError, "`weights` must be a DataArray"):
data.weighted([1, 2])
@pytest.mark.parametrize("as_dataset", (True, False))
@pytest.mark.parametrize("weights", ([np.nan, 2], [np.nan, np.nan]))
def test_weighted_weights_nan_raises(as_dataset, weights):
data = DataArray([1, 2])
if as_dataset:
data = data.to_dataset(name="data")
with pytest.raises(ValueError, match="`weights` cannot contain missing values."):
data.weighted(DataArray(weights))
@requires_dask
@pytest.mark.parametrize("as_dataset", (True, False))
@pytest.mark.parametrize("weights", ([np.nan, 2], [np.nan, np.nan]))
def test_weighted_weights_nan_raises_dask(as_dataset, weights):
data = DataArray([1, 2]).chunk({"dim_0": -1})
if as_dataset:
data = data.to_dataset(name="data")
weights = DataArray(weights).chunk({"dim_0": -1})
with raise_if_dask_computes():
weighted = data.weighted(weights)
with pytest.raises(ValueError, match="`weights` cannot contain missing values."):
weighted.sum().load()
@pytest.mark.parametrize(
("weights", "expected"),
(([1, 2], 3), ([2, 0], 2), ([0, 0], np.nan), ([-1, 1], np.nan)),
)
def test_weighted_sum_of_weights_no_nan(weights, expected):
da = DataArray([1, 2])
weights = DataArray(weights)
result = da.weighted(weights).sum_of_weights()
expected = DataArray(expected)
assert_equal(expected, result)
@pytest.mark.parametrize(
("weights", "expected"),
(([1, 2], 2), ([2, 0], np.nan), ([0, 0], np.nan), ([-1, 1], 1)),
)
def test_weighted_sum_of_weights_nan(weights, expected):
da = DataArray([np.nan, 2])
weights = DataArray(weights)
result = da.weighted(weights).sum_of_weights()
expected = DataArray(expected)
assert_equal(expected, result)
def test_weighted_sum_of_weights_bool():
# https://github.com/pydata/xarray/issues/4074
da = DataArray([1, 2])
weights = DataArray([True, True])
result = da.weighted(weights).sum_of_weights()
expected = DataArray(2)
assert_equal(expected, result)
@pytest.mark.parametrize("da", ([1.0, 2], [1, np.nan], [np.nan, np.nan]))
@pytest.mark.parametrize("factor", [0, 1, 3.14])
@pytest.mark.parametrize("skipna", (True, False))
def test_weighted_sum_equal_weights(da, factor, skipna):
# if all weights are 'f'; weighted sum is f times the ordinary sum
da = DataArray(da)
weights = xr.full_like(da, factor)
expected = da.sum(skipna=skipna) * factor
result = da.weighted(weights).sum(skipna=skipna)
assert_equal(expected, result)
@pytest.mark.parametrize(
("weights", "expected"), (([1, 2], 5), ([0, 2], 4), ([0, 0], 0))
)
def test_weighted_sum_no_nan(weights, expected):
da = DataArray([1, 2])
weights = DataArray(weights)
result = da.weighted(weights).sum()
expected = DataArray(expected)
assert_equal(expected, result)
@pytest.mark.parametrize(
("weights", "expected"), (([1, 2], 4), ([0, 2], 4), ([1, 0], 0), ([0, 0], 0))
)
@pytest.mark.parametrize("skipna", (True, False))
def test_weighted_sum_nan(weights, expected, skipna):
da = DataArray([np.nan, 2])
weights = DataArray(weights)
result = da.weighted(weights).sum(skipna=skipna)
if skipna:
expected = DataArray(expected)
else:
expected = DataArray(np.nan)
assert_equal(expected, result)
@pytest.mark.filterwarnings("error")
@pytest.mark.parametrize("da", ([1.0, 2], [1, np.nan], [np.nan, np.nan]))
@pytest.mark.parametrize("skipna", (True, False))
@pytest.mark.parametrize("factor", [1, 2, 3.14])
def test_weighted_mean_equal_weights(da, skipna, factor):
# if all weights are equal (!= 0), should yield the same result as mean
da = DataArray(da)
# all weights as 1.
weights = xr.full_like(da, factor)
expected = da.mean(skipna=skipna)
result = da.weighted(weights).mean(skipna=skipna)
assert_equal(expected, result)
@pytest.mark.parametrize(
("weights", "expected"), (([4, 6], 1.6), ([1, 0], 1.0), ([0, 0], np.nan))
)
def test_weighted_mean_no_nan(weights, expected):
da = DataArray([1, 2])
weights = DataArray(weights)
expected = DataArray(expected)
result = da.weighted(weights).mean()
assert_equal(expected, result)
@pytest.mark.parametrize(
("weights", "expected"), (([4, 6], 2.0), ([1, 0], np.nan), ([0, 0], np.nan))
)
@pytest.mark.parametrize("skipna", (True, False))
def test_weighted_mean_nan(weights, expected, skipna):
da = DataArray([np.nan, 2])
weights = DataArray(weights)
if skipna:
expected = DataArray(expected)
else:
expected = DataArray(np.nan)
result = da.weighted(weights).mean(skipna=skipna)
assert_equal(expected, result)
def test_weighted_mean_bool():
# https://github.com/pydata/xarray/issues/4074
da = DataArray([1, 1])
weights = DataArray([True, True])
expected = DataArray(1)
result = da.weighted(weights).mean()
assert_equal(expected, result)
def expected_weighted(da, weights, dim, skipna, operation):
"""
Generate expected result using ``*`` and ``sum``. This is checked against
the result of da.weighted which uses ``dot``
"""
weighted_sum = (da * weights).sum(dim=dim, skipna=skipna)
if operation == "sum":
return weighted_sum
masked_weights = weights.where(da.notnull())
sum_of_weights = masked_weights.sum(dim=dim, skipna=True)
valid_weights = sum_of_weights != 0
sum_of_weights = sum_of_weights.where(valid_weights)
if operation == "sum_of_weights":
return sum_of_weights
weighted_mean = weighted_sum / sum_of_weights
if operation == "mean":
return weighted_mean
@pytest.mark.parametrize("dim", ("a", "b", "c", ("a", "b"), ("a", "b", "c"), None))
@pytest.mark.parametrize("operation", ("sum_of_weights", "sum", "mean"))
@pytest.mark.parametrize("add_nans", (True, False))
@pytest.mark.parametrize("skipna", (None, True, False))
@pytest.mark.parametrize("as_dataset", (True, False))
def test_weighted_operations_3D(dim, operation, add_nans, skipna, as_dataset):
dims = ("a", "b", "c")
coords = dict(a=[0, 1, 2, 3], b=[0, 1, 2, 3], c=[0, 1, 2, 3])
weights = DataArray(np.random.randn(4, 4, 4), dims=dims, coords=coords)
data = np.random.randn(4, 4, 4)
# add approximately 25 % NaNs (https://stackoverflow.com/a/32182680/3010700)
if add_nans:
c = int(data.size * 0.25)
data.ravel()[np.random.choice(data.size, c, replace=False)] = np.NaN
data = DataArray(data, dims=dims, coords=coords)
if as_dataset:
data = data.to_dataset(name="data")
if operation == "sum_of_weights":
result = data.weighted(weights).sum_of_weights(dim)
else:
result = getattr(data.weighted(weights), operation)(dim, skipna=skipna)
expected = expected_weighted(data, weights, dim, skipna, operation)
assert_allclose(expected, result)
@pytest.mark.parametrize("operation", ("sum_of_weights", "sum", "mean"))
@pytest.mark.parametrize("as_dataset", (True, False))
def test_weighted_operations_nonequal_coords(operation, as_dataset):
weights = DataArray(np.random.randn(4), dims=("a",), coords=dict(a=[0, 1, 2, 3]))
data = DataArray(np.random.randn(4), dims=("a",), coords=dict(a=[1, 2, 3, 4]))
if as_dataset:
data = data.to_dataset(name="data")
expected = expected_weighted(
data, weights, dim="a", skipna=None, operation=operation
)
result = getattr(data.weighted(weights), operation)(dim="a")
assert_allclose(expected, result)
@pytest.mark.parametrize("dim", ("dim_0", None))
@pytest.mark.parametrize("shape_data", ((4,), (4, 4), (4, 4, 4)))
@pytest.mark.parametrize("shape_weights", ((4,), (4, 4), (4, 4, 4)))
@pytest.mark.parametrize("operation", ("sum_of_weights", "sum", "mean"))
@pytest.mark.parametrize("add_nans", (True, False))
@pytest.mark.parametrize("skipna", (None, True, False))
@pytest.mark.parametrize("as_dataset", (True, False))
def test_weighted_operations_different_shapes(
dim, shape_data, shape_weights, operation, add_nans, skipna, as_dataset
):
weights = DataArray(np.random.randn(*shape_weights))
data = np.random.randn(*shape_data)
# add approximately 25 % NaNs
if add_nans:
c = int(data.size * 0.25)
data.ravel()[np.random.choice(data.size, c, replace=False)] = np.NaN
data = DataArray(data)
if as_dataset:
data = data.to_dataset(name="data")
if operation == "sum_of_weights":
result = getattr(data.weighted(weights), operation)(dim)
else:
result = getattr(data.weighted(weights), operation)(dim, skipna=skipna)
expected = expected_weighted(data, weights, dim, skipna, operation)
assert_allclose(expected, result)
@pytest.mark.parametrize("operation", ("sum_of_weights", "sum", "mean"))
@pytest.mark.parametrize("as_dataset", (True, False))
@pytest.mark.parametrize("keep_attrs", (True, False, None))
def test_weighted_operations_keep_attr(operation, as_dataset, keep_attrs):
weights = DataArray(np.random.randn(2, 2), attrs=dict(attr="weights"))
data = DataArray(np.random.randn(2, 2))
if as_dataset:
data = data.to_dataset(name="data")
data.attrs = dict(attr="weights")
result = getattr(data.weighted(weights), operation)(keep_attrs=True)
if operation == "sum_of_weights":
assert weights.attrs == result.attrs
else:
assert data.attrs == result.attrs
result = getattr(data.weighted(weights), operation)(keep_attrs=None)
assert not result.attrs
result = getattr(data.weighted(weights), operation)(keep_attrs=False)
assert not result.attrs
@pytest.mark.parametrize("operation", ("sum", "mean"))
def test_weighted_operations_keep_attr_da_in_ds(operation):
# GH #3595
weights = DataArray(np.random.randn(2, 2))
data = DataArray(np.random.randn(2, 2), attrs=dict(attr="data"))
data = data.to_dataset(name="a")
result = getattr(data.weighted(weights), operation)(keep_attrs=True)
assert data.a.attrs == result.a.attrs
|
