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
|
# -*- coding: utf-8 -*-
import operator
from cpython cimport (PyObject_RichCompareBool,
Py_EQ, Py_NE, Py_LT, Py_LE, Py_GT, Py_GE)
import cython
from cython import Py_ssize_t
import numpy as np
from numpy cimport ndarray, uint8_t, import_array
import_array()
from pandas._libs.util cimport UINT8_MAX, is_nan
from pandas._libs.missing cimport checknull
@cython.wraparound(False)
@cython.boundscheck(False)
def scalar_compare(object[:] values, object val, object op):
"""
Compare each element of `values` array with the scalar `val`, with
the comparison operation described by `op`.
Parameters
----------
values : ndarray[object]
val : object
op : {operator.eq, operator.ne,
operator.le, operator.lt,
operator.ge, operator.gt}
Returns
-------
result : ndarray[bool]
"""
cdef:
Py_ssize_t i, n = len(values)
ndarray[uint8_t, cast=True] result
bint isnull_val
int flag
object x
if op is operator.lt:
flag = Py_LT
elif op is operator.le:
flag = Py_LE
elif op is operator.gt:
flag = Py_GT
elif op is operator.ge:
flag = Py_GE
elif op is operator.eq:
flag = Py_EQ
elif op is operator.ne:
flag = Py_NE
else:
raise ValueError('Unrecognized operator')
result = np.empty(n, dtype=bool).view(np.uint8)
isnull_val = checknull(val)
if flag == Py_NE:
for i in range(n):
x = values[i]
if checknull(x):
result[i] = True
elif isnull_val:
result[i] = True
else:
try:
result[i] = PyObject_RichCompareBool(x, val, flag)
except TypeError:
result[i] = True
elif flag == Py_EQ:
for i in range(n):
x = values[i]
if checknull(x):
result[i] = False
elif isnull_val:
result[i] = False
else:
try:
result[i] = PyObject_RichCompareBool(x, val, flag)
except TypeError:
result[i] = False
else:
for i in range(n):
x = values[i]
if checknull(x):
result[i] = False
elif isnull_val:
result[i] = False
else:
result[i] = PyObject_RichCompareBool(x, val, flag)
return result.view(bool)
@cython.wraparound(False)
@cython.boundscheck(False)
def vec_compare(object[:] left, object[:] right, object op):
"""
Compare the elements of `left` with the elements of `right` pointwise,
with the comparison operation described by `op`.
Parameters
----------
left : ndarray[object]
right : ndarray[object]
op : {operator.eq, operator.ne,
operator.le, operator.lt,
operator.ge, operator.gt}
Returns
-------
result : ndarray[bool]
"""
cdef:
Py_ssize_t i, n = len(left)
ndarray[uint8_t, cast=True] result
int flag
if n != len(right):
raise ValueError('Arrays were different lengths: {n} vs {nright}'
.format(n=n, nright=len(right)))
if op is operator.lt:
flag = Py_LT
elif op is operator.le:
flag = Py_LE
elif op is operator.gt:
flag = Py_GT
elif op is operator.ge:
flag = Py_GE
elif op is operator.eq:
flag = Py_EQ
elif op is operator.ne:
flag = Py_NE
else:
raise ValueError('Unrecognized operator')
result = np.empty(n, dtype=bool).view(np.uint8)
if flag == Py_NE:
for i in range(n):
x = left[i]
y = right[i]
if checknull(x) or checknull(y):
result[i] = True
else:
result[i] = PyObject_RichCompareBool(x, y, flag)
else:
for i in range(n):
x = left[i]
y = right[i]
if checknull(x) or checknull(y):
result[i] = False
else:
result[i] = PyObject_RichCompareBool(x, y, flag)
return result.view(bool)
@cython.wraparound(False)
@cython.boundscheck(False)
def scalar_binop(object[:] values, object val, object op):
"""
Apply the given binary operator `op` between each element of the array
`values` and the scalar `val`.
Parameters
----------
values : ndarray[object]
val : object
op : binary operator
Returns
-------
result : ndarray[object]
"""
cdef:
Py_ssize_t i, n = len(values)
object[:] result
object x
result = np.empty(n, dtype=object)
if val is None or is_nan(val):
result[:] = val
return result.base # `.base` to access underlying np.ndarray
for i in range(n):
x = values[i]
if x is None or is_nan(x):
result[i] = x
else:
result[i] = op(x, val)
return maybe_convert_bool(result.base)
@cython.wraparound(False)
@cython.boundscheck(False)
def vec_binop(object[:] left, object[:] right, object op):
"""
Apply the given binary operator `op` pointwise to the elements of
arrays `left` and `right`.
Parameters
----------
left : ndarray[object]
right : ndarray[object]
op : binary operator
Returns
-------
result : ndarray[object]
"""
cdef:
Py_ssize_t i, n = len(left)
object[:] result
if n != len(right):
raise ValueError('Arrays were different lengths: {n} vs {nright}'
.format(n=n, nright=len(right)))
result = np.empty(n, dtype=object)
for i in range(n):
x = left[i]
y = right[i]
try:
result[i] = op(x, y)
except TypeError:
if x is None or is_nan(x):
result[i] = x
elif y is None or is_nan(y):
result[i] = y
else:
raise
return maybe_convert_bool(result.base) # `.base` to access np.ndarray
def maybe_convert_bool(ndarray[object] arr,
true_values=None, false_values=None):
cdef:
Py_ssize_t i, n
ndarray[uint8_t] result
object val
set true_vals, false_vals
int na_count = 0
n = len(arr)
result = np.empty(n, dtype=np.uint8)
# the defaults
true_vals = {'True', 'TRUE', 'true'}
false_vals = {'False', 'FALSE', 'false'}
if true_values is not None:
true_vals = true_vals | set(true_values)
if false_values is not None:
false_vals = false_vals | set(false_values)
for i in range(n):
val = arr[i]
if isinstance(val, bool):
if val is True:
result[i] = 1
else:
result[i] = 0
elif val in true_vals:
result[i] = 1
elif val in false_vals:
result[i] = 0
elif isinstance(val, float):
result[i] = UINT8_MAX
na_count += 1
else:
return arr
if na_count > 0:
mask = result == UINT8_MAX
arr = result.view(np.bool_).astype(object)
np.putmask(arr, mask, np.nan)
return arr
else:
return result.view(np.bool_)
|
