/
sputils.py
388 lines (313 loc) · 12 KB
/
sputils.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
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
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
""" Utility functions for sparse matrix module
"""
from __future__ import division, print_function, absolute_import
import warnings
import numpy as np
__all__ = ['upcast', 'getdtype', 'isscalarlike', 'isintlike',
'isshape', 'issequence', 'isdense', 'ismatrix', 'get_sum_dtype']
supported_dtypes = ['bool', 'int8', 'uint8', 'short', 'ushort', 'intc',
'uintc', 'longlong', 'ulonglong', 'single', 'double',
'longdouble', 'csingle', 'cdouble', 'clongdouble']
supported_dtypes = [np.typeDict[x] for x in supported_dtypes]
_upcast_memo = {}
def upcast(*args):
"""Returns the nearest supported sparse dtype for the
combination of one or more types.
upcast(t0, t1, ..., tn) -> T where T is a supported dtype
Examples
--------
>>> upcast('int32')
<type 'numpy.int32'>
>>> upcast('bool')
<type 'numpy.bool_'>
>>> upcast('int32','float32')
<type 'numpy.float64'>
>>> upcast('bool',complex,float)
<type 'numpy.complex128'>
"""
t = _upcast_memo.get(hash(args))
if t is not None:
return t
upcast = np.find_common_type(args, [])
for t in supported_dtypes:
if np.can_cast(upcast, t):
_upcast_memo[hash(args)] = t
return t
raise TypeError('no supported conversion for types: %r' % (args,))
def upcast_char(*args):
"""Same as `upcast` but taking dtype.char as input (faster)."""
t = _upcast_memo.get(args)
if t is not None:
return t
t = upcast(*map(np.dtype, args))
_upcast_memo[args] = t
return t
def upcast_scalar(dtype, scalar):
"""Determine data type for binary operation between an array of
type `dtype` and a scalar.
"""
return (np.array([0], dtype=dtype) * scalar).dtype
def downcast_intp_index(arr):
"""
Down-cast index array to np.intp dtype if it is of a larger dtype.
Raise an error if the array contains a value that is too large for
intp.
"""
if arr.dtype.itemsize > np.dtype(np.intp).itemsize:
if arr.size == 0:
return arr.astype(np.intp)
maxval = arr.max()
minval = arr.min()
if maxval > np.iinfo(np.intp).max or minval < np.iinfo(np.intp).min:
raise ValueError("Cannot deal with arrays with indices larger "
"than the machine maximum address size "
"(e.g. 64-bit indices on 32-bit machine).")
return arr.astype(np.intp)
return arr
def to_native(A):
return np.asarray(A, dtype=A.dtype.newbyteorder('native'))
def getdtype(dtype, a=None, default=None):
"""Function used to simplify argument processing. If 'dtype' is not
specified (is None), returns a.dtype; otherwise returns a np.dtype
object created from the specified dtype argument. If 'dtype' and 'a'
are both None, construct a data type out of the 'default' parameter.
Furthermore, 'dtype' must be in 'allowed' set.
"""
# TODO is this really what we want?
if dtype is None:
try:
newdtype = a.dtype
except AttributeError:
if default is not None:
newdtype = np.dtype(default)
else:
raise TypeError("could not interpret data type")
else:
newdtype = np.dtype(dtype)
if newdtype == np.object_:
warnings.warn("object dtype is not supported by sparse matrices")
return newdtype
def get_index_dtype(arrays=(), maxval=None, check_contents=False):
"""
Based on input (integer) arrays `a`, determine a suitable index data
type that can hold the data in the arrays.
Parameters
----------
arrays : tuple of array_like
Input arrays whose types/contents to check
maxval : float, optional
Maximum value needed
check_contents : bool, optional
Whether to check the values in the arrays and not just their types.
Default: False (check only the types)
Returns
-------
dtype : dtype
Suitable index data type (int32 or int64)
"""
int32max = np.iinfo(np.int32).max
dtype = np.intc
if maxval is not None:
if maxval > int32max:
dtype = np.int64
if isinstance(arrays, np.ndarray):
arrays = (arrays,)
for arr in arrays:
arr = np.asarray(arr)
if arr.dtype > np.int32:
if check_contents:
if arr.size == 0:
# a bigger type not needed
continue
elif np.issubdtype(arr.dtype, np.integer):
maxval = arr.max()
minval = arr.min()
if (minval >= np.iinfo(np.int32).min and
maxval <= np.iinfo(np.int32).max):
# a bigger type not needed
continue
dtype = np.int64
break
return dtype
def get_sum_dtype(dtype):
"""Mimic numpy's casting for np.sum"""
if np.issubdtype(dtype, np.float_):
return np.float_
if dtype.kind == 'u' and np.can_cast(dtype, np.uint):
return np.uint
if np.can_cast(dtype, np.int_):
return np.int_
return dtype
def isscalarlike(x):
"""Is x either a scalar, an array scalar, or a 0-dim array?"""
return np.isscalar(x) or (isdense(x) and x.ndim == 0)
def isintlike(x):
"""Is x appropriate as an index into a sparse matrix? Returns True
if it can be cast safely to a machine int.
"""
if issequence(x):
return False
try:
return bool(int(x) == x)
except (TypeError, ValueError):
return False
def isshape(x):
"""Is x a valid 2-tuple of dimensions?
"""
try:
# Assume it's a tuple of matrix dimensions (M, N)
(M, N) = x
except:
return False
else:
if isintlike(M) and isintlike(N):
if np.ndim(M) == 0 and np.ndim(N) == 0:
return True
return False
def issequence(t):
return ((isinstance(t, (list, tuple)) and
(len(t) == 0 or np.isscalar(t[0]))) or
(isinstance(t, np.ndarray) and (t.ndim == 1)))
def ismatrix(t):
return ((isinstance(t, (list, tuple)) and
len(t) > 0 and issequence(t[0])) or
(isinstance(t, np.ndarray) and t.ndim == 2))
def isdense(x):
return isinstance(x, np.ndarray)
def validateaxis(axis):
if axis is not None:
axis_type = type(axis)
# In NumPy, you can pass in tuples for 'axis', but they are
# not very useful for sparse matrices given their limited
# dimensions, so let's make it explicit that they are not
# allowed to be passed in
if axis_type == tuple:
raise TypeError(("Tuples are not accepted for the 'axis' "
"parameter. Please pass in one of the "
"following: {-2, -1, 0, 1, None}."))
# If not a tuple, check that the provided axis is actually
# an integer and raise a TypeError similar to NumPy's
if not np.issubdtype(np.dtype(axis_type), np.integer):
raise TypeError("axis must be an integer, not {name}"
.format(name=axis_type.__name__))
if not (-2 <= axis <= 1):
raise ValueError("axis out of range")
class IndexMixin(object):
"""
This class simply exists to hold the methods necessary for fancy indexing.
"""
def _slicetoarange(self, j, shape):
""" Given a slice object, use numpy arange to change it to a 1D
array.
"""
start, stop, step = j.indices(shape)
return np.arange(start, stop, step)
def _unpack_index(self, index):
""" Parse index. Always return a tuple of the form (row, col).
Where row/col is a integer, slice, or array of integers.
"""
# First, check if indexing with single boolean matrix.
from .base import spmatrix # This feels dirty but...
if (isinstance(index, (spmatrix, np.ndarray)) and
(index.ndim == 2) and index.dtype.kind == 'b'):
return index.nonzero()
# Parse any ellipses.
index = self._check_ellipsis(index)
# Next, parse the tuple or object
if isinstance(index, tuple):
if len(index) == 2:
row, col = index
elif len(index) == 1:
row, col = index[0], slice(None)
else:
raise IndexError('invalid number of indices')
else:
row, col = index, slice(None)
# Next, check for validity, or transform the index as needed.
row, col = self._check_boolean(row, col)
return row, col
def _check_ellipsis(self, index):
"""Process indices with Ellipsis. Returns modified index."""
if index is Ellipsis:
return (slice(None), slice(None))
elif isinstance(index, tuple):
# Find first ellipsis
for j, v in enumerate(index):
if v is Ellipsis:
first_ellipsis = j
break
else:
first_ellipsis = None
# Expand the first one
if first_ellipsis is not None:
# Shortcuts
if len(index) == 1:
return (slice(None), slice(None))
elif len(index) == 2:
if first_ellipsis == 0:
if index[1] is Ellipsis:
return (slice(None), slice(None))
else:
return (slice(None), index[1])
else:
return (index[0], slice(None))
# General case
tail = ()
for v in index[first_ellipsis+1:]:
if v is not Ellipsis:
tail = tail + (v,)
nd = first_ellipsis + len(tail)
nslice = max(0, 2 - nd)
return index[:first_ellipsis] + (slice(None),)*nslice + tail
return index
def _check_boolean(self, row, col):
from .base import isspmatrix # ew...
# Supporting sparse boolean indexing with both row and col does
# not work because spmatrix.ndim is always 2.
if isspmatrix(row) or isspmatrix(col):
raise IndexError(
"Indexing with sparse matrices is not supported "
"except boolean indexing where matrix and index "
"are equal shapes.")
if isinstance(row, np.ndarray) and row.dtype.kind == 'b':
row = self._boolean_index_to_array(row)
if isinstance(col, np.ndarray) and col.dtype.kind == 'b':
col = self._boolean_index_to_array(col)
return row, col
def _boolean_index_to_array(self, i):
if i.ndim > 1:
raise IndexError('invalid index shape')
return i.nonzero()[0]
def _index_to_arrays(self, i, j):
i, j = self._check_boolean(i, j)
i_slice = isinstance(i, slice)
if i_slice:
i = self._slicetoarange(i, self.shape[0])[:, None]
else:
i = np.atleast_1d(i)
if isinstance(j, slice):
j = self._slicetoarange(j, self.shape[1])[None, :]
if i.ndim == 1:
i = i[:, None]
elif not i_slice:
raise IndexError('index returns 3-dim structure')
elif isscalarlike(j):
# row vector special case
j = np.atleast_1d(j)
if i.ndim == 1:
i, j = np.broadcast_arrays(i, j)
i = i[:, None]
j = j[:, None]
return i, j
else:
j = np.atleast_1d(j)
if i_slice and j.ndim > 1:
raise IndexError('index returns 3-dim structure')
i, j = np.broadcast_arrays(i, j)
if i.ndim == 1:
# return column vectors for 1-D indexing
i = i[None, :]
j = j[None, :]
elif i.ndim > 2:
raise IndexError("Index dimension must be <= 2")
return i, j