10BC0 MAINT Consistent cython types from _typedefs (#25942) · Veghit/scikit-learn@9b801fc · GitHub
[go: up one dir, main page]
Skip to content

Commit 9b801fc

Browse files
jeremiedbbjjerphan
authored and
Itay
committed
MAINT Consistent cython types from _typedefs (scikit-learn#25942)
Co-authored-by: Julien Jerphanion <git@jjerphan.xyz>
1 parent 5325d9c commit 9b801fc

File tree

12 files changed

+898
-922
lines changed

12 files changed

+898
-922
lines changed

‎setup.py‎

Lines changed: 3 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -265,7 +265,6 @@ def check_package_status(package, min_version):
265265
{
266266
"sources": ["_middle_term_computer.pyx.tp", "_middle_term_computer.pxd.tp"],
267267
"language": "c++",
268-
"include_np": True,
269268
"extra_compile_args": ["-std=c++11"],
270269
},
271270
{
@@ -397,9 +396,9 @@ def check_package_status(package, min_version):
397396
},
398397
{"sources": ["_random.pyx"], "include_np": True},
399398
{"sources": ["_logistic_sigmoid.pyx"], "include_np": True},
400-
{"sources": ["_typedefs.pyx"], "include_np": True},
401-
{"sources": ["_heap.pyx"], "include_np": True},
402-
{"sources": ["_sorting.pyx"], "include_np": True},
399+
{"sources": ["_typedefs.pyx"]},
400+
{"sources": ["_heap.pyx"]},
401+
{"sources": ["_sorting.pyx"]},
403402
{"sources": ["_vector_sentinel.pyx"], "language": "c++", "include_np": True},
404403
{"sources": ["_isfinite.pyx"]},
405404
],

‎sklearn/cluster/_hierarchical_fast.pyx‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,7 @@ cimport cython
55

66
from ..metrics._dist_metrics cimport DistanceMetric
77
from ..utils._fast_dict cimport IntFloatDict
8-
from ..utils._typedefs cimport float64_t, intp_t, bool_t
8+
from ..utils._typedefs cimport float64_t, intp_t, uint8_t
99

1010
# C++
1111
from cython.operator cimport dereference as deref, preincrement as inc
@@ -119,7 +119,7 @@ def _get_parents(
119119
nodes,
120120
heads,
121121
const intp_t[:] parents,
122-
bool_t[::1] not_visited
122+
uint8_t[::1] not_visited
123123
):
124124
"""Returns the heads of the given nodes, as defined by parents.
125125
@@ -465,7 +465,7 @@ def mst_linkage_core(
465465
"""
466466
cdef:
467467
intp_t n_samples = raw_data.shape[0]
468-
bool_t[:] in_tree = np.zeros(n_samples, dtype=bool)
468+
uint8_t[:] in_tree = np.zeros(n_samples, dtype=bool)
469469
float64_t[:, ::1] result = np.zeros((n_samples - 1 CB8A , 3))
470470

471471
intp_t current_node = 0

‎sklearn/metrics/_dist_metrics.pxd.tp‎

Lines changed: 54 additions & 59 deletions
Original file line numberDiff line numberDiff line change
@@ -9,26 +9,21 @@ implementation_specific_values = [
99
# for the float64 case as to still be able to expose the original
1010
# float64 implementation under the same API, namely `DistanceMetric`.
1111
#
12-
# On the other hand, '32' bit is used for `name_suffix` for the float32
12+
# On the other hand, '32' is used for `name_suffix` for the float32
1313
# case to remove ambiguity and use `DistanceMetric32`, which is not
1414
# publicly exposed.
1515
#
1616
# The metric mapping is adapted accordingly to route to the correct
1717
# implementations.
1818
#
19-
# We also use 64bit types as defined in `sklearn.utils._typedefs`
20-
# to maintain backward compatibility at the symbol level for extra
21-
# safety.
22-
#
23-
('', 'DTYPE_t', 'DTYPE'),
24-
('32', 'cnp.float32_t', 'np.float32')
19+
('', 'float64_t', 'np.float64'),
20+
('32', 'float32_t', 'np.float32')
2521
]
2622

2723
}}
28-
cimport numpy as cnp
2924
from libc.math cimport sqrt, exp
3025

31-
from ..utils._typedefs cimport DTYPE_t, ITYPE_t, SPARSE_INDEX_TYPE_t
26+
from ..utils._typedefs cimport float64_t, float32_t, int32_t, intp_t
3227

3328
{{for name_suffix, INPUT_DTYPE_t, INPUT_DTYPE in implementation_specific_values}}
3429

@@ -37,37 +32,37 @@ from ..utils._typedefs cimport DTYPE_t, ITYPE_t, SPARSE_INDEX_TYPE_t
3732
#
3833
# We use these for the default (euclidean) case so that they can be
3934
# inlined. This leads to faster computation for the most common case
40-
cdef inline DTYPE_t euclidean_dist{{name_suffix}}(
35+
cdef inline float64_t euclidean_dist{{name_suffix}}(
4136
const {{INPUT_DTYPE_t}}* x1,
4237
const {{INPUT_DTYPE_t}}* x2,
43-
ITYPE_t size,
38+
intp_t size,
4439
) except -1 nogil:
45-
cdef DTYPE_t tmp, d=0
46-
cdef cnp.intp_t j
40+
cdef float64_t tmp, d=0
41+
cdef intp_t j
4742
for j in range(size):
48-
tmp = <DTYPE_t> (x1[j] - x2[j])
43+
tmp = <float64_t> (x1[j] - x2[j])
4944
d += tmp * tmp
5045
return sqrt(d)
5146

5247

53-
cdef inline DTYPE_t euclidean_rdist{{name_suffix}}(
48+
cdef inline float64_t euclidean_rdist{{name_suffix}}(
5449
const {{INPUT_DTYPE_t}}* x1,
5550
const {{INPUT_DTYPE_t}}* x2,
56-
ITYPE_t size,
51+
intp_t size,
5752
) except -1 nogil:
58-
cdef DTYPE_t tmp, d=0
59-
cdef cnp.intp_t j
53+
cdef float64_t tmp, d=0
54+
cdef intp_t j
6055
for j in range(size):
61-
tmp = <DTYPE_t>(x1[j] - x2[j])
56+
tmp = <float64_t>(x1[j] - x2[j])
6257
d += tmp * tmp
6358
return d
6459

6560

66-
cdef inline DTYPE_t euclidean_dist_to_rdist{{name_suffix}}(const {{INPUT_DTYPE_t}} dist) except -1 nogil:
61+
cdef inline float64_t euclidean_dist_to_rdist{{name_suffix}}(const {{INPUT_DTYPE_t}} dist) except -1 nogil:
6762
return dist * dist
6863

6964

70-
cdef inline DTYPE_t euclidean_rdist_to_dist{{name_suffix}}(const {{INPUT_DTYPE_t}} dist) except -1 nogil:
65+
cdef inline float64_t euclidean_rdist_to_dist{{name_suffix}}(const {{INPUT_DTYPE_t}} dist) except -1 nogil:
7166
return sqrt(dist)
7267

7368

@@ -78,89 +73,89 @@ cdef class DistanceMetric{{name_suffix}}:
7873
# we must define them here so that cython's limited polymorphism will work.
7974
# Because we don't expect to instantiate a lot of these objects, the
8075
# extra memory overhead of this setup should not be an issue.
81-
cdef DTYPE_t p
82-
cdef const DTYPE_t[::1] vec
83-
cdef const DTYPE_t[:, ::1] mat
84-
cdef ITYPE_t size
76+
cdef float64_t p
77+
cdef const float64_t[::1] vec
78+
cdef const float64_t[:, ::1] mat
79+
cdef intp_t size
8580
cdef object func
8681
cdef object kwargs
8782

88-
cdef DTYPE_t dist(
83+
cdef float64_t dist(
8984
self,
9085
const {{INPUT_DTYPE_t}}* x1,
9186
const {{INPUT_DTYPE_t}}* x2,
92-
ITYPE_t size,
87+
intp_t size,
9388
) except -1 nogil
9489

95-
cdef DTYPE_t rdist(
90+
cdef float64_t rdist(
9691
self,
9792
const {{INPUT_DTYPE_t}}* x1,
9893
const {{INPUT_DTYPE_t}}* x2,
99-
ITYPE_t size,
94+
intp_t size,
10095
) except -1 nogil
10196

102-
cdef DTYPE_t dist_csr(
97+
cdef float64_t dist_csr(
10398
self,
10499
const {{INPUT_DTYPE_t}}* x1_data,
105-
const SPARSE_INDEX_TYPE_t[:] x1_indices,
100+
const int32_t[:] x1_indices,
106101
const {{INPUT_DTYPE_t}}* x2_data,
107-
const SPARSE_INDEX_TYPE_t[:] x2_indices,
108-
const SPARSE_INDEX_TYPE_t x1_start,
109-
const SPARSE_INDEX_TYPE_t x1_end,
110-
const SPARSE_INDEX_TYPE_t x2_start,
111-
const SPARSE_INDEX_TYPE_t x2_end,
112-
const ITYPE_t size,
102+
const int32_t[:] x2_indices,
103+
const int32_t x1_start,
104+
const int32_t x1_end,
105+
const int32_t x2_start,
106+
const int32_t x2_end,
107+
const intp_t size,
113108
) except -1 nogil
114109

115-
cdef DTYPE_t rdist_csr(
110+
cdef float64_t rdist_csr(
116111
self,
117112
const {{INPUT_DTYPE_t}}* x1_data,
118-
const SPARSE_INDEX_TYPE_t[:] x1_indices,
113+
const int32_t[:] x1_indices,
119114
const {{INPUT_DTYPE_t}}* x2_data,
120-
const SPARSE_INDEX_TYPE_t[:] x2_indices,
121-
const SPARSE_INDEX_TYPE_t x1_start,
122-
const SPARSE_INDEX_TYPE_t x1_end,
123-
const SPARSE_INDEX_TYPE_t x2_start,
124-
const SPARSE_INDEX_TYPE_t x2_end,
125-
const ITYPE_t size,
115+
const int32_t[:] x2_indices,
116+
const int32_t x1_start,
117+
const int32_t x1_end,
118+
const int32_t x2_start,
119+
const int32_t x2_end,
120+
const intp_t size,
126121
) except -1 nogil
127122

128123
cdef int pdist(
129124
self,
130125
const {{INPUT_DTYPE_t}}[:, ::1] X,
131-
DTYPE_t[:, ::1] D,
126+
float64_t[:, ::1] D,
132127
) except -1
133128

134129
cdef int cdist(
135130
self,
136131
const {{INPUT_DTYPE_t}}[:, ::1] X,
137132
const {{INPUT_DTYPE_t}}[:, ::1] Y,
138-
DTYPE_t[:, ::1] D,
133+
float64_t[:, ::1] D,
139134
) except -1
140135

141136
cdef int pdist_csr(
142137
self,
143138
const {{INPUT_DTYPE_t}}* x1_data,
144-
const SPARSE_INDEX_TYPE_t[:] x1_indices,
145-
const SPARSE_INDEX_TYPE_t[:] x1_indptr,
146-
const ITYPE_t size,
147-
DTYPE_t[:, ::1] D,
139+
const int32_t[:] x1_indices,
140+
const int32_t[:] x1_indptr,
141+
const intp_t size,
142+
float64_t[:, ::1] D,
148143
) except -1 nogil
149144

150145
cdef int cdist_csr(
151146
self,
152147
const {{INPUT_DTYPE_t}}* x1_data,
153-
const SPARSE_INDEX_TYPE_t[:] x1_indices,
154-
const SPARSE_INDEX_TYPE_t[:] x1_indptr,
148+
const int32_t[:] x1_indices,
149+
const int32_t[:] x1_indptr,
155150
const {{INPUT_DTYPE_t}}* x2_data,
156-
const SPARSE_INDEX_TYPE_t[:] x2_indices,
157-
const SPARSE_INDEX_TYPE_t[:] x2_indptr,
158-
const ITYPE_t size,
159-
DTYPE_t[:, ::1] D,
151+
const int32_t[:] x2_indices,
152+
const int32_t[:] x2_indptr,
153+
const intp_t size,
154+
float64_t[:, ::1] D,
160155
) except -1 nogil
161156

162-
cdef DTYPE_t _rdist_to_dist(self, {{INPUT_DTYPE_t}} rdist) except -1 nogil
157+
cdef float64_t _rdist_to_dist(self, {{INPUT_DTYPE_t}} rdist) except -1 nogil
163158

164-
cdef DTYPE_t _dist_to_rdist(self, {{INPUT_DTYPE_t}} dist) except -1 nogil
159+
cdef float64_t _dist_to_rdist(self, {{INPUT_DTYPE_t}} dist) except -1 nogil
165160

166161
{{endfor}}

0 commit comments

Comments
 (0)
0