-
Notifications
You must be signed in to change notification settings - Fork 9
Expand file tree
/
Copy pathnep-0049-data-allocation-strategies.html
More file actions
898 lines (700 loc) · 71.9 KB
/
nep-0049-data-allocation-strategies.html
File metadata and controls
898 lines (700 loc) · 71.9 KB
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
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
<!DOCTYPE html>
<html lang="en" data-content_root="./" >
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="viewport" content="width=device-width, initial-scale=1" />
<title>NEP 49 — Data allocation strategies — NumPy Enhancement Proposals</title>
<script data-cfasync="false">
document.documentElement.dataset.mode = localStorage.getItem("mode") || "";
document.documentElement.dataset.theme = localStorage.getItem("theme") || "";
</script>
<!--
this give us a css class that will be invisible only if js is disabled
-->
<noscript>
<style>
.pst-js-only { display: none !important; }
</style>
</noscript>
<!-- Loaded before other Sphinx assets -->
<link href="_static/styles/theme.css?digest=55c26ed414f5f8a31ebb" rel="stylesheet" />
<link href="_static/styles/pydata-sphinx-theme.css?digest=55c26ed414f5f8a31ebb" rel="stylesheet" />
<link rel="stylesheet" type="text/css" href="_static/pygments.css?v=03e43079" />
<!-- So that users can add custom icons -->
<script defer src="_static/scripts/fontawesome.js?digest=55c26ed414f5f8a31ebb"></script>
<!-- Pre-loaded scripts that we'll load fully later -->
<link rel="preload" as="script" href="_static/scripts/bootstrap.js?digest=55c26ed414f5f8a31ebb" />
<link rel="preload" as="script" href="_static/scripts/pydata-sphinx-theme.js?digest=55c26ed414f5f8a31ebb" />
<script src="_static/documentation_options.js?v=7f41d439"></script>
<script src="_static/doctools.js?v=888ff710"></script>
<script src="_static/sphinx_highlight.js?v=dc90522c"></script>
<script>DOCUMENTATION_OPTIONS.pagename = 'nep-0049-data-allocation-strategies';</script>
<script>DOCUMENTATION_OPTIONS.search_as_you_type = false;</script>
<link rel="icon" href="_static/favicon.ico"/>
<link rel="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
<link rel="next" title="NEP 50 — Promotion rules for Python scalars" href="nep-0050-scalar-promotion.html" />
<link rel="prev" title="NEP 40 — Legacy datatype implementation in NumPy" href="nep-0040-legacy-datatype-impl.html" />
<meta name="viewport" content="width=device-width, initial-scale=1"/>
<meta name="docsearch:language" content="en"/>
<meta name="docsearch:version" content="" />
<meta name="docbuild:last-update" content="Apr 24, 2026"/>
<script src="_static/searchtools.js"></script>
<script src="_static/language_data.js"></script>
<script src="searchindex.js"></script>
</head>
<body data-default-mode="">
<div id="pst-skip-link" class="skip-link d-print-none"><a href="#main-content">Skip to main content</a></div>
<div id="pst-scroll-pixel-helper"></div>
<button type="button" class="btn rounded-pill" id="pst-back-to-top">
<i class="fa-solid fa-arrow-up"></i>Back to top</button>
<dialog id="pst-search-dialog">
<form class="bd-search d-flex align-items-center"
action="search.html"
method="get">
<i class="fa-solid fa-magnifying-glass"></i>
<input type="search"
class="form-control"
name="q"
placeholder="Search the docs ..."
aria-label="Search the docs ..."
autocomplete="off"
autocorrect="off"
autocapitalize="off"
spellcheck="false"/>
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd>K</kbd></span>
</form>
</dialog>
<div class="pst-async-banner-revealer d-none">
<aside id="bd-header-version-warning" class="d-none d-print-none" aria-label="Version warning"></aside>
</div>
<header id="pst-header" class="bd-header navbar navbar-expand-lg bd-navbar d-print-none">
<div class="bd-header__inner bd-page-width">
<button class="pst-navbar-icon sidebar-toggle primary-toggle" aria-label="Site navigation">
<span class="fa-solid fa-bars"></span>
</button>
<div class="col-lg-3 navbar-header-items__start">
<div class="navbar-item">
<a class="navbar-brand logo" href="content.html">
<img src="_static/numpylogo.svg" class="logo__image only-light" alt="NumPy Enhancement Proposals - Home"/>
<img src="_static/numpylogo_dark.svg" class="logo__image only-dark pst-js-only" alt="NumPy Enhancement Proposals - Home"/>
</a></div>
</div>
<div class="col-lg-9 navbar-header-items">
<div class="me-auto navbar-header-items__center">
<div class="navbar-item">
<nav>
<ul class="bd-navbar-elements navbar-nav">
<li class="nav-item current active">
<a class="nav-link nav-internal" href="index.html">
Index
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="scope.html">
The Scope of NumPy
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="roadmap.html">
Current roadmap
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-external" href="https://github.com/numpy/numpy/issues?q=is%3Aopen+is%3Aissue+label%3A%2223+-+Wish+List%22">
Wishlist
</a>
</li>
</ul>
</nav></div>
</div>
<div class="navbar-header-items__end">
<div class="navbar-item navbar-persistent--container">
<button class="btn search-button-field search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="fa-solid fa-magnifying-glass"></i>
<span class="search-button__default-text">Search</span>
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd class="kbd-shortcut__modifier">K</kbd></span>
</button>
</div>
<div class="navbar-item">
<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="theme-switch fa-solid fa-sun fa-lg" data-mode="light" title="Light"></i>
<i class="theme-switch fa-solid fa-moon fa-lg" data-mode="dark" title="Dark"></i>
<i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto" title="System Settings"></i>
</button></div>
<div class="navbar-item"><ul class="navbar-icon-links"
aria-label="Icon Links">
<li class="nav-item">
<a href="https://github.com/numpy/numpy" title="GitHub" class="nav-link pst-navbar-icon" rel="noopener" target="_blank" data-bs-toggle="tooltip" data-bs-placement="bottom"><i class="fa-brands fa-square-github fa-lg" aria-hidden="true"></i>
<span class="sr-only">GitHub</span></a>
</li>
</ul></div>
</div>
</div>
<div class="navbar-persistent--mobile">
<button class="btn search-button-field search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="fa-solid fa-magnifying-glass"></i>
<span class="search-button__default-text">Search</span>
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd class="kbd-shortcut__modifier">K</kbd></span>
</button>
</div>
<button class="pst-navbar-icon sidebar-toggle secondary-toggle" aria-label="On this page">
<span class="fa-solid fa-outdent"></span>
</button>
</div>
</header>
<div class="bd-container">
<div class="bd-container__inner bd-page-width">
<dialog id="pst-primary-sidebar-modal"></dialog>
<div id="pst-primary-sidebar" class="bd-sidebar-primary bd-sidebar">
<div class="sidebar-header-items sidebar-primary__section">
<div class="sidebar-header-items__center">
<div class="navbar-item">
<nav>
<ul class="bd-navbar-elements navbar-nav">
<li class="nav-item current active">
<a class="nav-link nav-internal" href="index.html">
Index
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="scope.html">
The Scope of NumPy
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="roadmap.html">
Current roadmap
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-external" href="https://github.com/numpy/numpy/issues?q=is%3Aopen+is%3Aissue+label%3A%2223+-+Wish+List%22">
Wishlist
</a>
</li>
</ul>
</nav></div>
</div>
<div class="sidebar-header-items__end">
<div class="navbar-item">
<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="theme-switch fa-solid fa-sun fa-lg" data-mode="light" title="Light"></i>
<i class="theme-switch fa-solid fa-moon fa-lg" data-mode="dark" title="Dark"></i>
<i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto" title="System Settings"></i>
</button></div>
<div class="navbar-item"><ul class="navbar-icon-links"
aria-label="Icon Links">
<li class="nav-item">
<a href="https://github.com/numpy/numpy" title="GitHub" class="nav-link pst-navbar-icon" rel="noopener" target="_blank" data-bs-toggle="tooltip" data-bs-placement="bottom"><i class="fa-brands fa-square-github fa-lg" aria-hidden="true"></i>
<span class="sr-only">GitHub</span></a>
</li>
</ul></div>
</div>
</div>
<div class="sidebar-primary-items__start sidebar-primary__section">
<div class="sidebar-primary-item pst-sidebar-collapse"><button id="pst-collapse-sidebar-button" aria-expanded="true" aria-controls="pst-primary-sidebar"><svg class="pst-icon svg-inline--fa" role="img" aria-hidden="true" focusable="false" viewBox="0 0 16 16" xmlns="http://www.w3.org/2000/svg">
<path fill="currentColor" d="M3 15.5C2.36232 15.5 1.74874 15.2564 1.28478 14.8189C0.820828 14.3815 0.541576 13.7832 0.504167 13.1467L0.5 13L0.5 3C0.499965 2.36232 0.743605 1.74874 1.18107 1.28478C1.61854 0.820828 2.21676 0.541576 2.85333 0.504167L3 0.5L13 0.5C13.6377 0.499965 14.2513 0.743605 14.7152 1.18107C15.1792 1.61854 15.4584 2.21676 15.4958 2.85333L15.5 3L15.5 13C15.5 13.6377 15.2564 14.2513 14.8189 14.7152C14.3815 15.1792 13.7832 15.4584 13.1467 15.4958L13 15.5L3 15.5ZM3 13.8333L10.5 13.8333L10.5 2.16667L3 2.16667C2.79589 2.16669 2.59889 2.24163 2.44636 2.37726C2.29383 2.5129 2.19638 2.69979 2.1725 2.9025L2.16667 3L2.16667 13C2.16669 13.2041 2.24163 13.4011 2.37726 13.5536C2.5129 13.7062 2.69979 13.8036 2.9025 13.8275L3 13.8333ZM6.65583 10.325L6.5775 10.2558L4.91083 8.58917C4.76735 8.44567 4.68116 8.25476 4.66843 8.05223C4.65569 7.84971 4.71729 7.6495 4.84167 7.48917L4.91083 7.41083L6.5775 5.74417C6.72747 5.59471 6.9287 5.50794 7.14032 5.50148C7.35194 5.49502 7.55809 5.56935 7.7169 5.70937C7.8757 5.8494 7.97525 6.04463 7.99533 6.25539C8.01541 6.46616 7.95451 6.67667 7.825 6.84417L7.75583 6.9225L6.67917 8L7.75583 9.0775C7.89931 9.22099 7.98551 9.41191 7.99824 9.61443C8.01097 9.81695 7.94938 10.0172 7.825 10.1775L7.75583 10.2558C7.61234 10.3993 7.42142 10.4855 7.2189 10.4982C7.01638 10.511 6.81617 10.4494 6.65583 10.325Z"/>
</svg>
<span class="pst-collapse-sidebar-label">Collapse Sidebar</span>
<span class="pst-expand-sidebar-label">Expand Sidebar</span>
</button></div>
<div class="sidebar-primary-item">
<nav class="bd-docs-nav bd-links"
aria-label="Section Navigation">
<p class="bd-links__title" role="heading" aria-level="1">Section Navigation</p>
<div class="bd-toc-item navbar-nav"><ul class="nav bd-sidenav">
<li class="toctree-l1"><a class="reference internal" href="scope.html">The Scope of NumPy</a></li>
<li class="toctree-l1"><a class="reference internal" href="roadmap.html">Current roadmap</a></li>
</ul>
<ul class="current nav bd-sidenav">
<li class="toctree-l1 has-children"><a class="reference internal" href="meta.html">Meta-NEPs (NEPs about NEPs or active Processes)</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="nep-0000.html">NEP 0 — Purpose and process</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0023-backwards-compatibility.html">NEP 23 — Backwards compatibility and deprecation policy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0036-fair-play.html">NEP 36 — Fair play</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0045-c_style_guide.html">NEP 45 — C style guide</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0046-sponsorship-guidelines.html">NEP 46 — NumPy sponsorship guidelines</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0048-spending-project-funds.html">NEP 48 — Spending NumPy project funds</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-template.html">NEP X — Template and instructions</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="provisional.html">Provisional NEPs (provisionally accepted; interface may change)</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul class="simple">
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="accepted.html">Accepted NEPs (implementation in progress)</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="nep-0041-improved-dtype-support.html">NEP 41 — First step towards a new datatype system</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0042-new-dtypes.html">NEP 42 — New and extensible DTypes</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0044-restructuring-numpy-docs.html">NEP 44 — Restructuring the NumPy documentation</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0051-scalar-representation.html">NEP 51 — Changing the representation of NumPy scalars</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0054-simd-cpp-highway.html">NEP 54 — SIMD infrastructure evolution: adopting Google Highway when moving to C++</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="open.html">Open NEPs (under consideration)</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="nep-0043-extensible-ufuncs.html">NEP 43 — Enhancing the extensibility of UFuncs</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0053-c-abi-evolution.html">NEP 53 — Evolving the NumPy C-API for NumPy 2.0</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0057-numpy-platform-support.html">NEP 57 — NumPy platform support</a></li>
</ul>
</details></li>
<li class="toctree-l1 current active has-children"><a class="reference internal" href="finished.html">Finished NEPs</a><details open="open"><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul class="current">
<li class="toctree-l2"><a class="reference internal" href="nep-0001-npy-format.html">NEP 1 — A simple file format for NumPy arrays</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0005-generalized-ufuncs.html">NEP 5 — Generalized universal functions</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0007-datetime-proposal.html">NEP 7 — A proposal for implementing some date/time types in NumPy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0010-new-iterator-ufunc.html">NEP 10 — Optimizing iterator/UFunc performance</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0013-ufunc-overrides.html">NEP 13 — A mechanism for overriding Ufuncs</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0014-dropping-python2.7-proposal.html">NEP 14 — Plan for dropping Python 2.7 support</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0015-merge-multiarray-umath.html">NEP 15 — Merging multiarray and umath</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0018-array-function-protocol.html">NEP 18 — A dispatch mechanism for NumPy's high level array functions</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0019-rng-policy.html">NEP 19 — Random number generator policy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0020-gufunc-signature-enhancement.html">NEP 20 — Expansion of generalized universal function signatures</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0022-ndarray-duck-typing-overview.html">NEP 22 — Duck typing for NumPy arrays – high level overview</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0027-zero-rank-arrays.html">NEP 27 — Zero rank arrays</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0028-website-redesign.html">NEP 28 — numpy.org website redesign</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0029-deprecation_policy.html">NEP 29 — Recommend Python and NumPy version support as a community policy standard</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0032-remove-financial-functions.html">NEP 32 — Remove the financial functions from NumPy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0034-infer-dtype-is-object.html">NEP 34 — Disallow inferring ``dtype=object`` from sequences</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0035-array-creation-dispatch-with-array-function.html">NEP 35 — Array creation dispatching with __array_function__</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0038-SIMD-optimizations.html">NEP 38 — Using SIMD optimization instructions for performance</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0040-legacy-datatype-impl.html">NEP 40 — Legacy datatype implementation in NumPy</a></li>
<li class="toctree-l2 current active"><a class="current reference internal" href="#">NEP 49 — Data allocation strategies</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0050-scalar-promotion.html">NEP 50 — Promotion rules for Python scalars</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0052-python-api-cleanup.html">NEP 52 — Python API cleanup for NumPy 2.0</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0055-string_dtype.html">NEP 55 — Add a UTF-8 variable-width string DType to NumPy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0056-array-api-main-namespace.html">NEP 56 — Array API standard support in NumPy's main namespace</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="deferred.html">Deferred and Superseded NEPs</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="nep-0002-warnfix.html">NEP 2 — A proposal to build numpy without warning with a big set of warning flags</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0003-math_config_clean.html">NEP 3 — Cleaning the math configuration of numpy.core</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0004-datetime-proposal3.html">NEP 4 — A (third) proposal for implementing some date/time types in NumPy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0006-newbugtracker.html">NEP 6 — Replacing Trac with a different bug tracker</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0008-groupby_additions.html">NEP 8 — A proposal for adding groupby functionality to NumPy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0009-structured_array_extensions.html">NEP 9 — Structured array extensions</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0011-deferred-ufunc-evaluation.html">NEP 11 — Deferred UFunc evaluation</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0012-missing-data.html">NEP 12 — Missing data functionality in NumPy</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0021-advanced-indexing.html">NEP 21 — Simplified and explicit advanced indexing</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0024-missing-data-2.html">NEP 24 — Missing data functionality - alternative 1 to NEP 12</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0025-missing-data-3.html">NEP 25 — NA support via special dtypes</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0026-missing-data-summary.html">NEP 26 — Summary of missing data NEPs and discussion</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0030-duck-array-protocol.html">NEP 30 — Duck typing for NumPy arrays - implementation</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0031-uarray.html">NEP 31 — Context-local and global overrides of the NumPy API</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0037-array-module.html">NEP 37 — A dispatch protocol for NumPy-like modules</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0047-array-api-standard.html">NEP 47 — Adopting the array API standard</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="rejected.html">Rejected and Withdrawn NEPs</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="nep-0016-abstract-array.html">NEP 16 — An abstract base class for identifying "duck arrays"</a></li>
<li class="toctree-l2"><a class="reference internal" href="nep-0017-split-out-maskedarray.html">NEP 17 — Split out masked arrays</a></li>
</ul>
</details></li>
</ul>
</div>
</nav></div>
</div>
<div class="sidebar-primary-items__end sidebar-primary__section">
<div class="sidebar-primary-item">
<div id="ethical-ad-placement"
class="flat"
data-ea-publisher="readthedocs"
data-ea-type="readthedocs-sidebar"
data-ea-manual="true">
</div></div>
</div>
</div>
<main id="main-content" class="bd-main" role="main">
<div class="bd-content">
<div class="bd-article-container">
<div class="bd-header-article d-print-none">
<div class="header-article-items header-article__inner">
<div class="header-article-items__start">
<div class="header-article-item">
<nav aria-label="Breadcrumb" class="d-print-none">
<ul class="bd-breadcrumbs">
<li class="breadcrumb-item breadcrumb-home">
<a href="content.html" class="nav-link" aria-label="Home">
<i class="fa-solid fa-home"></i>
</a>
</li>
<li class="breadcrumb-item"><a href="index.html" class="nav-link">Roadmap & NumPy enhancement proposals</a></li>
<li class="breadcrumb-item"><a href="finished.html" class="nav-link">Finished NEPs</a></li>
<li class="breadcrumb-item active" aria-current="page"><span class="ellipsis">NEP 49 — Data allocation strategies</span></li>
</ul>
</nav>
</div>
</div>
</div>
</div>
<div id="searchbox"></div>
<article class="bd-article">
<section id="nep-49-data-allocation-strategies">
<span id="nep49"></span><h1>NEP 49 — Data allocation strategies<a class="headerlink" href="#nep-49-data-allocation-strategies" title="Link to this heading">#</a></h1>
<dl class="field-list simple">
<dt class="field-odd">Author<span class="colon">:</span></dt>
<dd class="field-odd"><p>Matti Picus</p>
</dd>
<dt class="field-even">Status<span class="colon">:</span></dt>
<dd class="field-even"><p>Final</p>
</dd>
<dt class="field-odd">Type<span class="colon">:</span></dt>
<dd class="field-odd"><p>Standards Track</p>
</dd>
<dt class="field-even">Created<span class="colon">:</span></dt>
<dd class="field-even"><p>2021-04-18</p>
</dd>
<dt class="field-odd">Resolution<span class="colon">:</span></dt>
<dd class="field-odd"><p><a class="reference external" href="https://mail.python.org/archives/list/numpy-discussion@python.org/thread/YZ3PNTXZUT27B6ITFAD3WRSM3T3SRVK4/#PKYXCTG4R5Q6LIRZC4SEWLNBM6GLRF26">NumPy Discussion</a></p>
</dd>
</dl>
<section id="abstract">
<h2>Abstract<a class="headerlink" href="#abstract" title="Link to this heading">#</a></h2>
<p>The <code class="docutils literal notranslate"><span class="pre">numpy.ndarray</span></code> requires additional memory allocations
to hold <code class="docutils literal notranslate"><span class="pre">numpy.ndarray.strides</span></code>, <code class="docutils literal notranslate"><span class="pre">numpy.ndarray.shape</span></code> and
<code class="docutils literal notranslate"><span class="pre">numpy.ndarray.data</span></code> attributes. These attributes are specially allocated
after creating the python object in <code class="docutils literal notranslate"><span class="pre">__new__</span></code> method.</p>
<p>This NEP proposes a mechanism to override the memory management strategy used
for <code class="docutils literal notranslate"><span class="pre">ndarray->data</span></code> with user-provided alternatives. This allocation holds
the data and can be very large. As accessing this data often becomes
a performance bottleneck, custom allocation strategies to guarantee data
alignment or pinning allocations to specialized memory hardware can enable
hardware-specific optimizations. The other allocations remain unchanged.</p>
</section>
<section id="motivation-and-scope">
<h2>Motivation and scope<a class="headerlink" href="#motivation-and-scope" title="Link to this heading">#</a></h2>
<p>Users may wish to override the internal data memory routines with ones of their
own. Two such use-cases are to ensure data alignment and to pin certain
allocations to certain NUMA cores. This desire for alignment was discussed
multiple times on the mailing list <a class="reference external" href="https://numpy-discussion.scipy.narkive.com/MvmMkJcK/numpy-arrays-data-allocation-and-simd-alignement">in 2005</a>, and in <a class="reference external" href="https://github.com/numpy/numpy/issues/5312">issue 5312</a> in 2014,
which led to <a class="reference external" href="https://github.com/numpy/numpy/pull/5457">PR 5457</a> and more mailing list discussions <a class="reference external" href="https://mail.python.org/archives/list/numpy-discussion@python.org/thread/YPC5BGPUMKT2MLBP6O3FMPC35LFM2CCH/#YPC5BGPUMKT2MLBP6O3FMPC35LFM2CCH">here</a> <a class="reference external" href="https://mail.python.org/archives/list/numpy-discussion@python.org/thread/IQK3EPIIRE3V4BPNAMJ2ZST3NUG2MK2A/#IQK3EPIIRE3V4BPNAMJ2ZST3NUG2MK2A">and here</a>. In
a comment on the issue <a class="reference external" href="https://github.com/numpy/numpy/issues/5312#issuecomment-315234656">from 2017</a>, a user described how 64-byte alignment
improved performance by 40x.</p>
<p>Also related is <a class="reference external" href="https://github.com/numpy/numpy/issues/14177">issue 14177</a> around the use of <code class="docutils literal notranslate"><span class="pre">madvise</span></code> and huge pages on
Linux.</p>
<p>Various tracing and profiling libraries like <a class="reference external" href="https://github.com/pythonspeed/filprofiler/blob/master/design/allocator-overrides.md">filprofiler</a> or <a class="reference external" href="https://github.com/boundarydevices/efence">electric fence</a>
override <code class="docutils literal notranslate"><span class="pre">malloc</span></code>.</p>
<p>The long CPython discussion of <a class="reference external" href="https://bugs.python.org/issue18835">BPO 18835</a> began with discussing the need for
<code class="docutils literal notranslate"><span class="pre">PyMem_Alloc32</span></code> and <code class="docutils literal notranslate"><span class="pre">PyMem_Alloc64</span></code>. The early conclusion was that the
cost (of wasted padding) vs. the benefit of aligned memory is best left to the
user, but then evolves into a discussion of various proposals to deal with
memory allocations, including <a class="reference external" href="https://www.python.org/dev/peps/pep-0445/">PEP 445</a> <a class="reference external" href="https://docs.python.org/3/c-api/memory.html#customize-memory-allocators">memory interfaces</a> to
<code class="docutils literal notranslate"><span class="pre">PyTraceMalloc_Track</span></code> which apparently was explicitly added for NumPy.</p>
<p>Allowing users to implement different strategies via the NumPy C-API will
enable exploration of this rich area of possible optimizations. The intention
is to create a flexible enough interface without burdening normative users.</p>
</section>
<section id="usage-and-impact">
<h2>Usage and impact<a class="headerlink" href="#usage-and-impact" title="Link to this heading">#</a></h2>
<p>The new functions can only be accessed via the NumPy C-API. An example is
included later in this NEP. The added <code class="docutils literal notranslate"><span class="pre">struct</span></code> will increase the size of the
<code class="docutils literal notranslate"><span class="pre">ndarray</span></code> object. It is a necessary price to pay for this approach. We
can be reasonably sure that the change in size will have a minimal impact on
end-user code because NumPy version 1.20 already changed the object size.</p>
<p>The implementation preserves the use of <code class="docutils literal notranslate"><span class="pre">PyTraceMalloc_Track</span></code> to track
allocations already present in NumPy.</p>
</section>
<section id="backward-compatibility">
<h2>Backward compatibility<a class="headerlink" href="#backward-compatibility" title="Link to this heading">#</a></h2>
<p>The design will not break backward compatibility. Projects that were assigning
to the <code class="docutils literal notranslate"><span class="pre">ndarray->data</span></code> pointer were already breaking the current memory
management strategy and should restore
<code class="docutils literal notranslate"><span class="pre">ndarray->data</span></code> before calling <code class="docutils literal notranslate"><span class="pre">Py_DECREF</span></code>. As mentioned above, the change
in size should not impact end-users.</p>
</section>
<section id="detailed-description">
<h2>Detailed description<a class="headerlink" href="#detailed-description" title="Link to this heading">#</a></h2>
<section id="high-level-design">
<h3>High level design<a class="headerlink" href="#high-level-design" title="Link to this heading">#</a></h3>
<p>Users who wish to change the NumPy data memory management routines will use
<a class="reference internal" href="#c.PyDataMem_SetHandler" title="PyDataMem_SetHandler"><code class="xref c c-func docutils literal notranslate"><span class="pre">PyDataMem_SetHandler()</span></code></a>, which uses a <a class="reference internal" href="#c.PyDataMem_Handler" title="PyDataMem_Handler"><code class="xref c c-type docutils literal notranslate"><span class="pre">PyDataMem_Handler</span></code></a>
structure to hold pointers to functions used to manage the data memory. In
order to allow lifetime management of the <code class="docutils literal notranslate"><span class="pre">context</span></code>, the structure is wrapped
in a <code class="docutils literal notranslate"><span class="pre">PyCapsule</span></code>.</p>
<p>Since a call to <code class="docutils literal notranslate"><span class="pre">PyDataMem_SetHandler</span></code> will change the default functions, but
that function may be called during the lifetime of an <code class="docutils literal notranslate"><span class="pre">ndarray</span></code> object, each
<code class="docutils literal notranslate"><span class="pre">ndarray</span></code> will carry with it the <code class="docutils literal notranslate"><span class="pre">PyDataMem_Handler</span></code>-wrapped PyCapsule used
at the time of its instantiation, and these will be used to reallocate or free
the data memory of the instance. Internally NumPy may use <code class="docutils literal notranslate"><span class="pre">memcpy</span></code> or
<code class="docutils literal notranslate"><span class="pre">memset</span></code> on the pointer to the data memory.</p>
<p>The name of the handler will be exposed on the python level via a
<code class="docutils literal notranslate"><span class="pre">numpy.core.multiarray.get_handler_name(arr)</span></code> function. If called as
<code class="docutils literal notranslate"><span class="pre">numpy.core.multiarray.get_handler_name()</span></code> it will return the name of the
handler that will be used to allocate data for the next new <cite>ndarrray</cite>.</p>
<p>The version of the handler will be exposed on the python level via a
<code class="docutils literal notranslate"><span class="pre">numpy.core.multiarray.get_handler_version(arr)</span></code> function. If called as
<code class="docutils literal notranslate"><span class="pre">numpy.core.multiarray.get_handler_version()</span></code> it will return the version of the
handler that will be used to allocate data for the next new <cite>ndarrray</cite>.</p>
<p>The version, currently 1, allows for future enhancements to the
<code class="docutils literal notranslate"><span class="pre">PyDataMemAllocator</span></code>. If fields are added, they must be added to the end.</p>
</section>
<section id="numpy-c-api-functions">
<h3>NumPy C-API functions<a class="headerlink" href="#numpy-c-api-functions" title="Link to this heading">#</a></h3>
<dl class="c type">
<dt class="sig sig-object c" id="c.PyDataMem_Handler">
<span class="k"><span class="pre">type</span></span><span class="w"> </span><span class="sig-name descname"><span class="n"><span class="pre">PyDataMem_Handler</span></span></span><a class="headerlink" href="#c.PyDataMem_Handler" title="Link to this definition">#</a><br /></dt>
<dd><p>A struct to hold function pointers used to manipulate memory</p>
<div class="highlight-c notranslate"><div class="highlight"><pre><span></span><span class="k">typedef</span><span class="w"> </span><span class="k">struct</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">char</span><span class="w"> </span><span class="n">name</span><span class="p">[</span><span class="mi">127</span><span class="p">];</span><span class="w"> </span><span class="cm">/* multiple of 64 to keep the struct aligned */</span>
<span class="w"> </span><span class="kt">uint8_t</span><span class="w"> </span><span class="n">version</span><span class="p">;</span><span class="w"> </span><span class="cm">/* currently 1 */</span>
<span class="w"> </span><span class="n">PyDataMemAllocator</span><span class="w"> </span><span class="n">allocator</span><span class="p">;</span>
<span class="p">}</span><span class="w"> </span><span class="n">PyDataMem_Handler</span><span class="p">;</span>
</pre></div>
</div>
<p>where the allocator structure is</p>
<div class="highlight-c notranslate"><div class="highlight"><pre><span></span><span class="cm">/* The declaration of free differs from PyMemAllocatorEx */</span>
<span class="k">typedef</span><span class="w"> </span><span class="k">struct</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">;</span>
<span class="w"> </span><span class="kt">void</span><span class="o">*</span><span class="w"> </span><span class="p">(</span><span class="o">*</span><span class="n">malloc</span><span class="p">)</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">size</span><span class="p">);</span>
<span class="w"> </span><span class="kt">void</span><span class="o">*</span><span class="w"> </span><span class="p">(</span><span class="o">*</span><span class="n">calloc</span><span class="p">)</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">nelem</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">elsize</span><span class="p">);</span>
<span class="w"> </span><span class="kt">void</span><span class="o">*</span><span class="w"> </span><span class="p">(</span><span class="o">*</span><span class="n">realloc</span><span class="p">)</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ptr</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">new_size</span><span class="p">);</span>
<span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="p">(</span><span class="o">*</span><span class="n">free</span><span class="p">)</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="n">ptr</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">size</span><span class="p">);</span>
<span class="p">}</span><span class="w"> </span><span class="n">PyDataMemAllocator</span><span class="p">;</span>
</pre></div>
</div>
<p>The use of a <code class="docutils literal notranslate"><span class="pre">size</span></code> parameter in <code class="docutils literal notranslate"><span class="pre">free</span></code> differentiates this struct from
the <a class="reference external" href="https://docs.python.org/dev/c-api/memory.html#c.PyMemAllocatorEx" title="(in Python v3.15)"><code class="xref c c-type docutils literal notranslate"><span class="pre">PyMemAllocatorEx</span></code></a> struct in Python. This call signature is
used internally in NumPy currently, and also in other places for instance
<cite>C++98 <https://en.cppreference.com/w/cpp/memory/allocator/deallocate></cite>,
<cite>C++11 <https://en.cppreference.com/w/cpp/memory/allocator_traits/deallocate></cite>, and
<cite>Rust (allocator_api) <https://doc.rust-lang.org/std/alloc/trait.Allocator.html#tymethod.deallocate></cite>.</p>
<p>The consumer of the <cite>PyDataMemAllocator</cite> interface must keep track of <code class="docutils literal notranslate"><span class="pre">size</span></code> and make sure it is
consistent with the parameter passed to the <code class="docutils literal notranslate"><span class="pre">(m|c|re)alloc</span></code> functions.</p>
<p>NumPy itself may violate this requirement when the shape of the requested
array contains a <code class="docutils literal notranslate"><span class="pre">0</span></code>, so authors of PyDataMemAllocators should relate to
the <code class="docutils literal notranslate"><span class="pre">size</span></code> parameter as a best-guess. Work to fix this is ongoing in PRs
<a class="reference external" href="https://github.com/numpy/numpy/pull/15780">15780</a> and <a class="reference external" href="https://github.com/numpy/numpy/pull/15788">15788</a> but has not yet been resolved. When it is this NEP should
be revisited.</p>
</dd></dl>
<dl class="c function">
<dt class="sig sig-object c" id="c.PyDataMem_SetHandler">
<a class="reference external" href="https://docs.python.org/dev/c-api/structures.html#c.PyObject" title="(in Python v3.15)"><span class="n"><span class="pre">PyObject</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">PyDataMem_SetHandler</span></span></span><span class="sig-paren">(</span><a class="reference external" href="https://docs.python.org/dev/c-api/structures.html#c.PyObject" title="(in Python v3.15)"><span class="n"><span class="pre">PyObject</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="n"><span class="pre">handler</span></span><span class="sig-paren">)</span><a class="headerlink" href="#c.PyDataMem_SetHandler" title="Link to this definition">#</a><br /></dt>
<dd><p>Sets a new allocation policy. If the input value is <code class="docutils literal notranslate"><span class="pre">NULL</span></code>, will reset
the policy to the default. Return the previous policy, or
return NULL if an error has occurred. We wrap the user-provided
so they will still call the Python and NumPy memory management callback
hooks. All the function pointers must be filled in, <code class="docutils literal notranslate"><span class="pre">NULL</span></code> is not
accepted.</p>
</dd></dl>
<dl class="c function">
<dt class="sig sig-object c" id="c.PyDataMem_GetHandler">
<span class="k"><span class="pre">const</span></span><span class="w"> </span><a class="reference external" href="https://docs.python.org/dev/c-api/structures.html#c.PyObject" title="(in Python v3.15)"><span class="n"><span class="pre">PyObject</span></span></a><span class="w"> </span><span class="p"><span class="pre">*</span></span><span class="sig-name descname"><span class="n"><span class="pre">PyDataMem_GetHandler</span></span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="headerlink" href="#c.PyDataMem_GetHandler" title="Link to this definition">#</a><br /></dt>
<dd><p>Return the current policy that will be used to allocate data for the
next <code class="docutils literal notranslate"><span class="pre">PyArrayObject</span></code>. On failure, return <code class="docutils literal notranslate"><span class="pre">NULL</span></code>.</p>
</dd></dl>
</section>
<section id="pydatamem-handler-thread-safety-and-lifetime">
<h3><code class="docutils literal notranslate"><span class="pre">PyDataMem_Handler</span></code> thread safety and lifetime<a class="headerlink" href="#pydatamem-handler-thread-safety-and-lifetime" title="Link to this heading">#</a></h3>
<p>The active handler is stored in the current <a class="reference external" href="https://docs.python.org/dev/library/contextvars.html#contextvars.Context" title="(in Python v3.15)"><code class="xref py py-class docutils literal notranslate"><span class="pre">Context</span></code></a>
via a <a class="reference external" href="https://docs.python.org/dev/library/contextvars.html#contextvars.ContextVar" title="(in Python v3.15)"><code class="xref py py-class docutils literal notranslate"><span class="pre">ContextVar</span></code></a>. This ensures it can be configured both
per-thread and per-async-coroutine.</p>
<p>There is currently no lifetime management of <code class="docutils literal notranslate"><span class="pre">PyDataMem_Handler</span></code>.
The user of <cite>PyDataMem_SetHandler</cite> must ensure that the argument remains alive
for as long as any objects allocated with it, and while it is the active handler.
In practice, this means the handler must be immortal.</p>
<p>As an implementation detail, currently this <code class="docutils literal notranslate"><span class="pre">ContextVar</span></code> contains a <code class="docutils literal notranslate"><span class="pre">PyCapsule</span></code>
object storing a pointer to a <code class="docutils literal notranslate"><span class="pre">PyDataMem_Handler</span></code> with no destructor,
but this should not be relied upon.</p>
</section>
<section id="sample-code">
<h3>Sample code<a class="headerlink" href="#sample-code" title="Link to this heading">#</a></h3>
<p>This code adds a 64-byte header to each <code class="docutils literal notranslate"><span class="pre">data</span></code> pointer and stores information
about the allocation in the header. Before calling <code class="docutils literal notranslate"><span class="pre">free</span></code>, a check ensures
the <code class="docutils literal notranslate"><span class="pre">sz</span></code> argument is correct.</p>
<div class="highlight-c notranslate"><div class="highlight"><pre><span></span><span class="cp">#define NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION</span>
<span class="cp">#include</span><span class="w"> </span><span class="cpf"><numpy/arrayobject.h></span>
<span class="n">NPY_NO_EXPORT</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span>
<span class="k">typedef</span><span class="w"> </span><span class="k">struct</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">(</span><span class="o">*</span><span class="n">malloc</span><span class="p">)(</span><span class="kt">size_t</span><span class="p">);</span>
<span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">(</span><span class="o">*</span><span class="n">calloc</span><span class="p">)(</span><span class="kt">size_t</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="p">);</span>
<span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">(</span><span class="o">*</span><span class="n">realloc</span><span class="p">)(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="p">);</span>
<span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="p">(</span><span class="o">*</span><span class="n">free</span><span class="p">)(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">);</span>
<span class="p">}</span><span class="w"> </span><span class="n">Allocator</span><span class="p">;</span>
<span class="n">NPY_NO_EXPORT</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span>
<span class="n">shift_alloc</span><span class="p">(</span><span class="n">Allocator</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">sz</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="n">real</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">(</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">ctx</span><span class="o">-></span><span class="n">malloc</span><span class="p">(</span><span class="n">sz</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">);</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">real</span><span class="w"> </span><span class="o">==</span><span class="w"> </span><span class="nb">NULL</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="nb">NULL</span><span class="p">;</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="n">snprintf</span><span class="p">(</span><span class="n">real</span><span class="p">,</span><span class="w"> </span><span class="mi">64</span><span class="p">,</span><span class="w"> </span><span class="s">"originally allocated %ld"</span><span class="p">,</span><span class="w"> </span><span class="p">(</span><span class="kt">unsigned</span><span class="w"> </span><span class="kt">long</span><span class="p">)</span><span class="n">sz</span><span class="p">);</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">)(</span><span class="n">real</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">);</span>
<span class="p">}</span>
<span class="n">NPY_NO_EXPORT</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span>
<span class="n">shift_zero</span><span class="p">(</span><span class="n">Allocator</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">sz</span><span class="p">,</span><span class="w"> </span><span class="kt">size_t</span><span class="w"> </span><span class="n">cnt</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="n">real</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">(</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">ctx</span><span class="o">-></span><span class="n">calloc</span><span class="p">(</span><span class="n">sz</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">,</span><span class="w"> </span><span class="n">cnt</span><span class="p">);</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">real</span><span class="w"> </span><span class="o">==</span><span class="w"> </span><span class="nb">NULL</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="nb">NULL</span><span class="p">;</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="n">snprintf</span><span class="p">(</span><span class="n">real</span><span class="p">,</span><span class="w"> </span><span class="mi">64</span><span class="p">,</span><span class="w"> </span><span class="s">"originally allocated %ld via zero"</span><span class="p">,</span>
<span class="w"> </span><span class="p">(</span><span class="kt">unsigned</span><span class="w"> </span><span class="kt">long</span><span class="p">)</span><span class="n">sz</span><span class="p">);</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">)(</span><span class="n">real</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">);</span>
<span class="p">}</span>
<span class="n">NPY_NO_EXPORT</span><span class="w"> </span><span class="kt">void</span>
<span class="n">shift_free</span><span class="p">(</span><span class="n">Allocator</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="w"> </span><span class="n">p</span><span class="p">,</span><span class="w"> </span><span class="n">npy_uintp</span><span class="w"> </span><span class="n">sz</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">p</span><span class="w"> </span><span class="o">==</span><span class="w"> </span><span class="nb">NULL</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="p">;</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="n">real</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">(</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">p</span><span class="w"> </span><span class="o">-</span><span class="w"> </span><span class="mi">64</span><span class="p">;</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">strncmp</span><span class="p">(</span><span class="n">real</span><span class="p">,</span><span class="w"> </span><span class="s">"originally allocated"</span><span class="p">,</span><span class="w"> </span><span class="mi">20</span><span class="p">)</span><span class="w"> </span><span class="o">!=</span><span class="w"> </span><span class="mi">0</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">fprintf</span><span class="p">(</span><span class="n">stdout</span><span class="p">,</span><span class="w"> </span><span class="s">"uh-oh, unmatched shift_free, "</span>
<span class="w"> </span><span class="s">"no appropriate prefix</span><span class="se">\\</span><span class="s">n"</span><span class="p">);</span>
<span class="w"> </span><span class="cm">/* Make C runtime crash by calling free on the wrong address */</span>
<span class="w"> </span><span class="n">ctx</span><span class="o">-></span><span class="n">free</span><span class="p">((</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">p</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">10</span><span class="p">);</span>
<span class="w"> </span><span class="cm">/* ctx->free(real); */</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="k">else</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">npy_uintp</span><span class="w"> </span><span class="n">i</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">(</span><span class="n">npy_uintp</span><span class="p">)</span><span class="n">atoi</span><span class="p">(</span><span class="n">real</span><span class="w"> </span><span class="o">+</span><span class="mi">20</span><span class="p">);</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">i</span><span class="w"> </span><span class="o">!=</span><span class="w"> </span><span class="n">sz</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">fprintf</span><span class="p">(</span><span class="n">stderr</span><span class="p">,</span><span class="w"> </span><span class="s">"uh-oh, unmatched shift_free"</span>
<span class="w"> </span><span class="s">"(ptr, %ld) but allocated %ld</span><span class="se">\\</span><span class="s">n"</span><span class="p">,</span><span class="w"> </span><span class="n">sz</span><span class="p">,</span><span class="w"> </span><span class="n">i</span><span class="p">);</span>
<span class="w"> </span><span class="cm">/* This happens when the shape has a 0, only print */</span>
<span class="w"> </span><span class="n">ctx</span><span class="o">-></span><span class="n">free</span><span class="p">(</span><span class="n">real</span><span class="p">);</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="k">else</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">ctx</span><span class="o">-></span><span class="n">free</span><span class="p">(</span><span class="n">real</span><span class="p">);</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="p">}</span>
<span class="p">}</span>
<span class="n">NPY_NO_EXPORT</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span>
<span class="n">shift_realloc</span><span class="p">(</span><span class="n">Allocator</span><span class="w"> </span><span class="o">*</span><span class="n">ctx</span><span class="p">,</span><span class="w"> </span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="w"> </span><span class="n">p</span><span class="p">,</span><span class="w"> </span><span class="n">npy_uintp</span><span class="w"> </span><span class="n">sz</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">p</span><span class="w"> </span><span class="o">!=</span><span class="w"> </span><span class="nb">NULL</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="n">real</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">(</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">p</span><span class="w"> </span><span class="o">-</span><span class="w"> </span><span class="mi">64</span><span class="p">;</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">strncmp</span><span class="p">(</span><span class="n">real</span><span class="p">,</span><span class="w"> </span><span class="s">"originally allocated"</span><span class="p">,</span><span class="w"> </span><span class="mi">20</span><span class="p">)</span><span class="w"> </span><span class="o">!=</span><span class="w"> </span><span class="mi">0</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">fprintf</span><span class="p">(</span><span class="n">stdout</span><span class="p">,</span><span class="w"> </span><span class="s">"uh-oh, unmatched shift_realloc</span><span class="se">\\</span><span class="s">n"</span><span class="p">);</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="n">realloc</span><span class="p">(</span><span class="n">p</span><span class="p">,</span><span class="w"> </span><span class="n">sz</span><span class="p">);</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">)((</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">ctx</span><span class="o">-></span><span class="n">realloc</span><span class="p">(</span><span class="n">real</span><span class="p">,</span><span class="w"> </span><span class="n">sz</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">)</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">);</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="k">else</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="n">real</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">(</span><span class="kt">char</span><span class="w"> </span><span class="o">*</span><span class="p">)</span><span class="n">ctx</span><span class="o">-></span><span class="n">realloc</span><span class="p">(</span><span class="n">p</span><span class="p">,</span><span class="w"> </span><span class="n">sz</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">);</span>
<span class="w"> </span><span class="k">if</span><span class="w"> </span><span class="p">(</span><span class="n">real</span><span class="w"> </span><span class="o">==</span><span class="w"> </span><span class="nb">NULL</span><span class="p">)</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="nb">NULL</span><span class="p">;</span>
<span class="w"> </span><span class="p">}</span>
<span class="w"> </span><span class="n">snprintf</span><span class="p">(</span><span class="n">real</span><span class="p">,</span><span class="w"> </span><span class="mi">64</span><span class="p">,</span><span class="w"> </span><span class="s">"originally allocated "</span>
<span class="w"> </span><span class="s">"%ld via realloc"</span><span class="p">,</span><span class="w"> </span><span class="p">(</span><span class="kt">unsigned</span><span class="w"> </span><span class="kt">long</span><span class="p">)</span><span class="n">sz</span><span class="p">);</span>
<span class="w"> </span><span class="k">return</span><span class="w"> </span><span class="p">(</span><span class="kt">void</span><span class="w"> </span><span class="o">*</span><span class="p">)(</span><span class="n">real</span><span class="w"> </span><span class="o">+</span><span class="w"> </span><span class="mi">64</span><span class="p">);</span>
<span class="w"> </span><span class="p">}</span>
<span class="p">}</span>
<span class="k">static</span><span class="w"> </span><span class="n">Allocator</span><span class="w"> </span><span class="n">new_handler_ctx</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="n">malloc</span><span class="p">,</span>
<span class="w"> </span><span class="n">calloc</span><span class="p">,</span>
<span class="w"> </span><span class="n">realloc</span><span class="p">,</span>
<span class="w"> </span><span class="n">free</span>
<span class="p">};</span>
<span class="k">static</span><span class="w"> </span><span class="n">PyDataMem_Handler</span><span class="w"> </span><span class="n">new_handler</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="s">"secret_data_allocator"</span><span class="p">,</span>
<span class="w"> </span><span class="mi">1</span><span class="p">,</span>
<span class="w"> </span><span class="p">{</span>
<span class="w"> </span><span class="o">&</span><span class="n">new_handler_ctx</span><span class="p">,</span>
<span class="w"> </span><span class="n">shift_alloc</span><span class="p">,</span><span class="w"> </span><span class="cm">/* malloc */</span>
<span class="w"> </span><span class="n">shift_zero</span><span class="p">,</span><span class="w"> </span><span class="cm">/* calloc */</span>
<span class="w"> </span><span class="n">shift_realloc</span><span class="p">,</span><span class="w"> </span><span class="cm">/* realloc */</span>
<span class="w"> </span><span class="n">shift_free</span><span class="w"> </span><span class="cm">/* free */</span>
<span class="w"> </span><span class="p">}</span>
<span class="p">};</span>
</pre></div>
</div>
</section>
</section>
<section id="related-work">
<h2>Related work<a class="headerlink" href="#related-work" title="Link to this heading">#</a></h2>
<p>This NEP is being tracked by the <a class="reference external" href="https://quansight.github.io/pnumpy/stable/index.html">pnumpy</a> project and a <a class="reference external" href="https://github.com/numpy/numpy/pull/17582#issuecomment-809145547">comment in the PR</a>
mentions use in orchestrating FPGA DMAs.</p>
</section>
<section id="implementation">
<h2>Implementation<a class="headerlink" href="#implementation" title="Link to this heading">#</a></h2>
<p>This NEP has been implemented in <a class="reference external" href="https://github.com/numpy/numpy/pull/17582">PR 17582</a>.</p>
</section>
<section id="alternatives">
<h2>Alternatives<a class="headerlink" href="#alternatives" title="Link to this heading">#</a></h2>
<p>These were discussed in <a class="reference external" href="https://github.com/numpy/numpy/issues/17467">issue 17467</a>. <a class="reference external" href="https://github.com/numpy/numpy/pull/5457">PR 5457</a> and <a class="reference external" href="https://github.com/numpy/numpy/pull/5470">PR 5470</a> proposed a
global interface for specifying aligned allocations.</p>
<p><code class="docutils literal notranslate"><span class="pre">PyArray_malloc_aligned</span></code> and friends were added to NumPy with the
<cite>numpy.random</cite> module API refactor. and are used there for performance.</p>
<p><a class="reference external" href="https://github.com/numpy/numpy/pull/390">PR 390</a> had two parts: expose <code class="docutils literal notranslate"><span class="pre">PyDataMem_*</span></code> via the NumPy C-API, and a hook
mechanism. The PR was merged with no example code for using these features.</p>
</section>
<section id="discussion">
<h2>Discussion<a class="headerlink" href="#discussion" title="Link to this heading">#</a></h2>
<p>The discussion on the mailing list led to the <code class="docutils literal notranslate"><span class="pre">PyDataMemAllocator</span></code> struct
with a <code class="docutils literal notranslate"><span class="pre">context</span></code> field like <a class="reference external" href="https://docs.python.org/dev/c-api/memory.html#c.PyMemAllocatorEx" title="(in Python v3.15)"><code class="xref c c-type docutils literal notranslate"><span class="pre">PyMemAllocatorEx</span></code></a> but with a different
signature for <code class="docutils literal notranslate"><span class="pre">free</span></code>.</p>
</section>
<section id="references-and-footnotes">
<h2>References and footnotes<a class="headerlink" href="#references-and-footnotes" title="Link to this heading">#</a></h2>
<aside class="footnote-list brackets">
<aside class="footnote brackets" id="id1" role="doc-footnote">
<span class="label"><span class="fn-bracket">[</span><a role="doc-backlink" href="#id4">1</a><span class="fn-bracket">]</span></span>
<p>Each NEP must either be explicitly labeled as placed in the public domain (see
this NEP as an example) or licensed under the <a class="reference external" href="https://www.opencontent.org/openpub/">Open Publication License</a>.</p>
</aside>
</aside>
</section>
<section id="copyright">
<h2>Copyright<a class="headerlink" href="#copyright" title="Link to this heading">#</a></h2>
<p>This document has been placed in the public domain. <a class="footnote-reference brackets" href="#id1" id="id4" role="doc-noteref"><span class="fn-bracket">[</span>1<span class="fn-bracket">]</span></a></p>
</section>
</section>
</article>
</div>
<dialog id="pst-secondary-sidebar-modal"></dialog>
<div id="pst-secondary-sidebar" class="bd-sidebar-secondary bd-toc"><div class="sidebar-secondary-items sidebar-secondary__inner">
<div class="sidebar-secondary-item">
<div
id="pst-page-navigation-heading-2"
class="page-toc tocsection onthispage">
<i class="fa-solid fa-list"></i> On this page
</div>
<nav id="pst-page-toc-nav" class="page-toc" aria-labelledby="pst-page-navigation-heading-2">
<ul class="pst-show_toc_level nav section-nav flex-column">
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#abstract">Abstract</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#motivation-and-scope">Motivation and scope</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#usage-and-impact">Usage and impact</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#backward-compatibility">Backward compatibility</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#detailed-description">Detailed description</a><ul class="nav section-nav flex-column">
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#high-level-design">High level design</a></li>
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#numpy-c-api-functions">NumPy C-API functions</a></li>
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#pydatamem-handler-thread-safety-and-lifetime"><code class="docutils literal notranslate"><span class="pre">PyDataMem_Handler</span></code> thread safety and lifetime</a></li>
<li class="toc-h3 nav-item toc-entry"><a class="reference internal nav-link" href="#sample-code">Sample code</a></li>
</ul>
</li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#related-work">Related work</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#implementation">Implementation</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#alternatives">Alternatives</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#discussion">Discussion</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#references-and-footnotes">References and footnotes</a></li>
<li class="toc-h2 nav-item toc-entry"><a class="reference internal nav-link" href="#copyright">Copyright</a></li>
</ul>
</nav></div>
</div></div>
</div>
<footer class="bd-footer-content">
</footer>
</main>
</div>
</div>
<!-- Scripts loaded after <body> so the DOM is not blocked -->
<script defer src="_static/scripts/bootstrap.js?digest=55c26ed414f5f8a31ebb"></script>
<script defer src="_static/scripts/pydata-sphinx-theme.js?digest=55c26ed414f5f8a31ebb"></script>
<footer class="bd-footer">
<div class="bd-footer__inner bd-page-width">
<div class="footer-items__start">
<div class="footer-item">
<p class="copyright">
© Copyright 2017-2026, NumPy Developers.
<br/>
</p>
</div>
<div class="footer-item">
<p class="sphinx-version">
Created using <a href="https://www.sphinx-doc.org/">Sphinx</a> 7.2.6.
<br/>
</p>
</div>
</div>
<div class="footer-items__end">
<div class="footer-item">
<p class="theme-version">
<!-- # L10n: Setting the PST URL as an argument as this does not need to be localized -->
Built with the <a href="https://pydata-sphinx-theme.readthedocs.io/en/stable/index.html">PyData Sphinx Theme</a> 0.17.1.
</p></div>
</div>
</div>
</footer>
</body>
</html>