forked from pytorch/pytorch.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathlinalg.html
2023 lines (1761 loc) · 197 KB
/
linalg.html
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
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
<meta name="robots" content="noindex">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>torch.linalg — PyTorch 1.8.1 documentation</title>
<link rel="canonical" href="https://pytorch.org/docs/stable/linalg.html"/>
<link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
<!-- <link rel="stylesheet" href="_static/pygments.css" type="text/css" /> -->
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/katex.min.css" type="text/css" />
<link rel="stylesheet" href="_static/css/jit.css" type="text/css" />
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/katex.min.css" type="text/css" />
<link rel="stylesheet" href="_static/katex-math.css" type="text/css" />
<link rel="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
<link rel="next" title="torch.overrides" href="torch.overrides.html" />
<link rel="prev" title="TorchScript Unsupported Pytorch Constructs" href="jit_unsupported.html" />
<script src="_static/js/modernizr.min.js"></script>
<!-- Preload the theme fonts -->
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-book.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/IBMPlexMono/IBMPlexMono-Medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-medium-italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/IBMPlexMono/IBMPlexMono-SemiBold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<!-- Preload the katex fonts -->
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Math-Italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Main-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Main-Bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size1-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size4-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size2-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size3-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Caligraphic-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.15.2/css/all.css" integrity="sha384-vSIIfh2YWi9wW0r9iZe7RJPrKwp6bG+s9QZMoITbCckVJqGCCRhc+ccxNcdpHuYu" crossorigin="anonymous">
</head>
<div class="container-fluid header-holder tutorials-header" id="header-holder">
<div class="container">
<div class="header-container">
<a class="header-logo" href="https://pytorch.org/" aria-label="PyTorch"></a>
<div class="main-menu">
<ul>
<li>
<a href="https://pytorch.org/get-started">Get Started</a>
</li>
<li>
<a href="https://pytorch.org/ecosystem">Ecosystem</a>
</li>
<li>
<a href="https://pytorch.org/mobile">Mobile</a>
</li>
<li>
<a href="https://pytorch.org/blog/">Blog</a>
</li>
<li>
<a href="https://pytorch.org/tutorials">Tutorials</a>
</li>
<li class="active">
<div id="resourcesDropdownButton" data-toggle="resources-dropdown" class="resources-dropdown">
<a class="resource-option with-down-orange-arrow">
Docs
</a>
<div class="resources-dropdown-menu">
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/docs/stable/index.html">
<span class="dropdown-title">PyTorch</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/audio/stable/index.html">
<span class="dropdown-title">torchaudio</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/text/stable/index.html">
<span class="dropdown-title">torchtext</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/vision/stable/index.html">
<span class="dropdown-title">torchvision</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/elastic/">
<span class="dropdown-title">TorchElastic</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/serve/">
<span class="dropdown-title">TorchServe</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/xla">
<span class="dropdown-title">PyTorch on XLA Devices</span>
<p></p>
</a>
</div>
</li>
<li>
<div id="resourcesDropdownButton" data-toggle="resources-dropdown" class="resources-dropdown">
<a class="resource-option with-down-arrow">
Resources
</a>
<div class="resources-dropdown-menu">
<a class="nav-dropdown-item" href="https://pytorch.org/features">
<span class="dropdown-title">About</span>
<p>Learn about PyTorch’s features and capabilities</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/#community-module">
<span class="dropdown-title">Community</span>
<p>Join the PyTorch developer community to contribute, learn, and get your questions answered.</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/resources">
<span class="dropdown-title">Developer Resources</span>
<p>Find resources and get questions answered</p>
</a>
<a class="nav-dropdown-item" href="https://discuss.pytorch.org/" target="_blank">
<span class="dropdown-title">Forums</span>
<p>A place to discuss PyTorch code, issues, install, research</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/hub">
<span class="dropdown-title">Models (Beta)</span>
<p>Discover, publish, and reuse pre-trained models</p>
</a>
</div>
</div>
</li>
<li>
<a href="https://github.com/pytorch/pytorch">Github</a>
</li>
</ul>
</div>
<a class="main-menu-open-button" href="#" data-behavior="open-mobile-menu"></a>
</div>
</div>
</div>
<body class="pytorch-body">
<div class="table-of-contents-link-wrapper">
<span>Table of Contents</span>
<a href="#" class="toggle-table-of-contents" data-behavior="toggle-table-of-contents"></a>
</div>
<nav data-toggle="wy-nav-shift" class="pytorch-left-menu" id="pytorch-left-menu">
<div class="pytorch-side-scroll">
<div class="pytorch-menu pytorch-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
<div class="pytorch-left-menu-search">
<div class="version">
<a href='https://pytorch.org/docs/versions.html'>1.8.1 ▼</a>
</div>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="search.html" method="get">
<input type="text" name="q" placeholder="Search Docs" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
</div>
<p class="caption"><span class="caption-text">Notes</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="notes/amp_examples.html">Automatic Mixed Precision examples</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/autograd.html">Autograd mechanics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/broadcasting.html">Broadcasting semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/cpu_threading_torchscript_inference.html">CPU threading and TorchScript inference</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/cuda.html">CUDA semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/ddp.html">Distributed Data Parallel</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/extending.html">Extending PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/faq.html">Frequently Asked Questions</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/large_scale_deployments.html">Features for large-scale deployments</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/modules.html">Modules</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/multiprocessing.html">Multiprocessing best practices</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/randomness.html">Reproducibility</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/serialization.html">Serialization semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/windows.html">Windows FAQ</a></li>
</ul>
<p class="caption"><span class="caption-text">Language Bindings</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="cpp_index.html">C++</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/javadoc/">Javadoc</a></li>
</ul>
<p class="caption"><span class="caption-text">Python API</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="torch.html">torch</a></li>
<li class="toctree-l1"><a class="reference internal" href="nn.html">torch.nn</a></li>
<li class="toctree-l1"><a class="reference internal" href="nn.functional.html">torch.nn.functional</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensors.html">torch.Tensor</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensor_attributes.html">Tensor Attributes</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensor_view.html">Tensor Views</a></li>
<li class="toctree-l1"><a class="reference internal" href="autograd.html">torch.autograd</a></li>
<li class="toctree-l1"><a class="reference internal" href="cuda.html">torch.cuda</a></li>
<li class="toctree-l1"><a class="reference internal" href="amp.html">torch.cuda.amp</a></li>
<li class="toctree-l1"><a class="reference internal" href="backends.html">torch.backends</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.html">torch.distributed</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributions.html">torch.distributions</a></li>
<li class="toctree-l1"><a class="reference internal" href="fft.html">torch.fft</a></li>
<li class="toctree-l1"><a class="reference internal" href="futures.html">torch.futures</a></li>
<li class="toctree-l1"><a class="reference internal" href="fx.html">torch.fx</a></li>
<li class="toctree-l1"><a class="reference internal" href="hub.html">torch.hub</a></li>
<li class="toctree-l1"><a class="reference internal" href="jit.html">torch.jit</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">torch.linalg</a></li>
<li class="toctree-l1"><a class="reference internal" href="torch.overrides.html">torch.overrides</a></li>
<li class="toctree-l1"><a class="reference internal" href="profiler.html">torch.profiler</a></li >
<li class="toctree-l1"><a class="reference internal" href="nn.init.html">torch.nn.init</a></li>
<li class="toctree-l1"><a class="reference internal" href="onnx.html">torch.onnx</a></li>
<li class="toctree-l1"><a class="reference internal" href="optim.html">torch.optim</a></li>
<li class="toctree-l1"><a class="reference internal" href="complex_numbers.html">Complex Numbers</a></li>
<li class="toctree-l1"><a class="reference internal" href="ddp_comm_hooks.html">DDP Communication Hooks</a></li>
<li class="toctree-l1"><a class="reference internal" href="pipeline.html">Pipeline Parallelism</a></li>
<li class="toctree-l1"><a class="reference internal" href="quantization.html">Quantization</a></li>
<li class="toctree-l1"><a class="reference internal" href="rpc.html">Distributed RPC Framework</a></li>
<li class="toctree-l1"><a class="reference internal" href="random.html">torch.random</a></li>
<li class="toctree-l1"><a class="reference internal" href="sparse.html">torch.sparse</a></li>
<li class="toctree-l1"><a class="reference internal" href="storage.html">torch.Storage</a></li>
<li class="toctree-l1"><a class="reference internal" href="benchmark_utils.html">torch.utils.benchmark</a></li>
<li class="toctree-l1"><a class="reference internal" href="bottleneck.html">torch.utils.bottleneck</a></li>
<li class="toctree-l1"><a class="reference internal" href="checkpoint.html">torch.utils.checkpoint</a></li>
<li class="toctree-l1"><a class="reference internal" href="cpp_extension.html">torch.utils.cpp_extension</a></li>
<li class="toctree-l1"><a class="reference internal" href="data.html">torch.utils.data</a></li>
<li class="toctree-l1"><a class="reference internal" href="dlpack.html">torch.utils.dlpack</a></li>
<li class="toctree-l1"><a class="reference internal" href="mobile_optimizer.html">torch.utils.mobile_optimizer</a></li>
<li class="toctree-l1"><a class="reference internal" href="model_zoo.html">torch.utils.model_zoo</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensorboard.html">torch.utils.tensorboard</a></li>
<li class="toctree-l1"><a class="reference internal" href="type_info.html">Type Info</a></li>
<li class="toctree-l1"><a class="reference internal" href="named_tensor.html">Named Tensors</a></li>
<li class="toctree-l1"><a class="reference internal" href="name_inference.html">Named Tensors operator coverage</a></li>
<li class="toctree-l1"><a class="reference internal" href="__config__.html">torch.__config__</a></li>
</ul>
<p class="caption"><span class="caption-text">Libraries</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/audio/stable">torchaudio</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/text/stable">torchtext</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/vision/stable">torchvision</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/elastic/">TorchElastic</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/serve">TorchServe</a></li>
<li class="toctree-l1"><a class="reference external" href="http://pytorch.org/xla/">PyTorch on XLA Devices</a></li>
</ul>
<p class="caption"><span class="caption-text">Community</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="community/contribution_guide.html">PyTorch Contribution Guide</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/governance.html">PyTorch Governance</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/persons_of_interest.html">PyTorch Governance | Persons of Interest</a></li>
</ul>
</div>
</div>
</nav>
<div class="pytorch-container">
<div class="pytorch-page-level-bar" id="pytorch-page-level-bar">
<div class="pytorch-breadcrumbs-wrapper">
<div role="navigation" aria-label="breadcrumbs navigation">
<ul class="pytorch-breadcrumbs">
<li>
<a href="index.html">
Docs
</a> >
</li>
<li>torch.linalg</li>
<li class="pytorch-breadcrumbs-aside">
<a href="_sources/linalg.rst.txt" rel="nofollow"><img src="_static/images/view-page-source-icon.svg"></a>
</li>
</ul>
</div>
</div>
<div class="pytorch-shortcuts-wrapper" id="pytorch-shortcuts-wrapper">
Shortcuts
</div>
</div>
<section data-toggle="wy-nav-shift" id="pytorch-content-wrap" class="pytorch-content-wrap">
<div class="pytorch-content-left">
<div class="rst-content">
<div role="main" class="main-content" itemscope="itemscope" itemtype="http://schema.org/Article">
<article itemprop="articleBody" id="pytorch-article" class="pytorch-article">
<div class="section" id="torch-linalg">
<h1>torch.linalg<a class="headerlink" href="#torch-linalg" title="Permalink to this headline">¶</a></h1>
<p>Common linear algebra operations.</p>
<p>This module is in BETA. New functions are still being added, and some
functions may change in future PyTorch releases. See the documentation of each
function for details.</p>
<span class="target" id="module-torch.linalg"></span><div class="section" id="functions">
<h2>Functions<a class="headerlink" href="#functions" title="Permalink to this headline">¶</a></h2>
<dl class="function">
<dt id="torch.linalg.cholesky">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">cholesky</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None</em><span class="sig-paren">)</span> → Tensor<a class="headerlink" href="#torch.linalg.cholesky" title="Permalink to this definition">¶</a></dt>
<dd><p>Computes the Cholesky decomposition of a Hermitian (or symmetric for real-valued matrices)
positive-definite matrix or the Cholesky decompositions for a batch of such matrices.
Each decomposition has the form:</p>
<div class="math">
<span class="katex-display"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><semantics><mrow><mtext>input</mtext><mo>=</mo><mi>L</mi><msup><mi>L</mi><mi>H</mi></msup></mrow><annotation encoding="application/x-tex">\text{input} = LL^H</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.8623000000000001em;vertical-align:-0.19444em;"></span><span class="mord text"><span class="mord">input</span></span><span class="mspace" style="margin-right:0.2777777777777778em;"></span><span class="mrel">=</span><span class="mspace" style="margin-right:0.2777777777777778em;"></span></span><span class="base"><span class="strut" style="height:0.8913309999999999em;vertical-align:0em;"></span><span class="mord mathnormal">L</span><span class="mord"><span class="mord mathnormal">L</span><span class="msupsub"><span class="vlist-t"><span class="vlist-r"><span class="vlist" style="height:0.8913309999999999em;"><span style="top:-3.113em;margin-right:0.05em;"><span class="pstrut" style="height:2.7em;"></span><span class="sizing reset-size6 size3 mtight"><span class="mord mathnormal mtight" style="margin-right:0.08125em;">H</span></span></span></span></span></span></span></span></span></span></span></span>
</div><p>where <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>L</mi></mrow><annotation encoding="application/x-tex">L</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.68333em;vertical-align:0em;"></span><span class="mord mathnormal">L</span></span></span></span>
</span> is a lower-triangular matrix and <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msup><mi>L</mi><mi>H</mi></msup></mrow><annotation encoding="application/x-tex">L^H</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.8413309999999999em;vertical-align:0em;"></span><span class="mord"><span class="mord mathnormal">L</span><span class="msupsub"><span class="vlist-t"><span class="vlist-r"><span class="vlist" style="height:0.8413309999999999em;"><span style="top:-3.063em;margin-right:0.05em;"><span class="pstrut" style="height:2.7em;"></span><span class="sizing reset-size6 size3 mtight"><span class="mord mathnormal mtight" style="margin-right:0.08125em;">H</span></span></span></span></span></span></span></span></span></span></span>
</span> is the conjugate transpose of <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>L</mi></mrow><annotation encoding="application/x-tex">L</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.68333em;vertical-align:0em;"></span><span class="mord mathnormal">L</span></span></span></span>
</span>,
which is just a transpose for the case of real-valued input matrices.
In code it translates to <code class="docutils literal notranslate"><span class="pre">input</span> <span class="pre">=</span> <span class="pre">L</span> <span class="pre">@</span> <span class="pre">L.t()</span></code> if <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is real-valued and
<code class="docutils literal notranslate"><span class="pre">input</span> <span class="pre">=</span> <span class="pre">L</span> <span class="pre">@</span> <span class="pre">L.conj().t()</span></code> if <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is complex-valued.
The batch of <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>L</mi></mrow><annotation encoding="application/x-tex">L</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.68333em;vertical-align:0em;"></span><span class="mord mathnormal">L</span></span></span></span>
</span> matrices is returned.</p>
<p>Supports real-valued and complex-valued inputs.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function synchronizes that device with the CPU.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>LAPACK’s <cite>potrf</cite> is used for CPU inputs, and MAGMA’s <cite>potrf</cite> is used for CUDA inputs.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>If <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is not a Hermitian positive-definite matrix, or if it’s a batch of matrices
and one or more of them is not a Hermitian positive-definite matrix, then a RuntimeError will be thrown.
If <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is a batch of matrices, then the error message will include the batch index
of the first matrix that is not Hermitian positive-definite.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the input tensor of size <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mo stretchy="false">(</mo><mo>∗</mo><mo separator="true">,</mo><mi>n</mi><mo separator="true">,</mo><mi>n</mi><mo stretchy="false">)</mo></mrow><annotation encoding="application/x-tex">(*, n, n)</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:1em;vertical-align:-0.25em;"></span><span class="mopen">(</span><span class="mord">∗</span><span class="mpunct">,</span><span class="mspace" style="margin-right:0.16666666666666666em;"></span><span class="mord mathnormal">n</span><span class="mpunct">,</span><span class="mspace" style="margin-right:0.16666666666666666em;"></span><span class="mord mathnormal">n</span><span class="mclose">)</span></span></span></span>
</span> consisting of Hermitian positive-definite
<span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>n</mi><mo>×</mo><mi>n</mi></mrow><annotation encoding="application/x-tex">n \times n</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.66666em;vertical-align:-0.08333em;"></span><span class="mord mathnormal">n</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span><span class="mbin">×</span><span class="mspace" style="margin-right:0.2222222222222222em;"></span></span><span class="base"><span class="strut" style="height:0.43056em;vertical-align:0em;"></span><span class="mord mathnormal">n</span></span></span></span>
</span> matrices, where <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mo>∗</mo></mrow><annotation encoding="application/x-tex">*</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.46528em;vertical-align:0em;"></span><span class="mord">∗</span></span></span></span>
</span> is zero or more batch dimensions.</p>
</dd>
<dt class="field-even">Keyword Arguments</dt>
<dd class="field-even"><p><strong>out</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a><em>, </em><em>optional</em>) – The output tensor. Ignored if <code class="docutils literal notranslate"><span class="pre">None</span></code>. Default: <code class="docutils literal notranslate"><span class="pre">None</span></code></p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex128</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">mm</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">a</span><span class="o">.</span><span class="n">t</span><span class="p">()</span><span class="o">.</span><span class="n">conj</span><span class="p">())</span> <span class="c1"># creates a Hermitian positive-definite matrix</span>
<span class="gp">>>> </span><span class="n">l</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cholesky</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[2.5266+0.0000j, 1.9586-2.0626j],</span>
<span class="go"> [1.9586+2.0626j, 9.4160+0.0000j]], dtype=torch.complex128)</span>
<span class="gp">>>> </span><span class="n">l</span>
<span class="go">tensor([[1.5895+0.0000j, 0.0000+0.0000j],</span>
<span class="go"> [1.2322+1.2976j, 2.4928+0.0000j]], dtype=torch.complex128)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">mm</span><span class="p">(</span><span class="n">l</span><span class="p">,</span> <span class="n">l</span><span class="o">.</span><span class="n">t</span><span class="p">()</span><span class="o">.</span><span class="n">conj</span><span class="p">())</span>
<span class="go">tensor([[2.5266+0.0000j, 1.9586-2.0626j],</span>
<span class="go"> [1.9586+2.0626j, 9.4160+0.0000j]], dtype=torch.complex128)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">float64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">a</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="o">-</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">))</span> <span class="c1"># creates a symmetric positive-definite matrix</span>
<span class="gp">>>> </span><span class="n">l</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cholesky</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[[ 1.1629, 2.0237],</span>
<span class="go"> [ 2.0237, 6.6593]],</span>
<span class="go"> [[ 0.4187, 0.1830],</span>
<span class="go"> [ 0.1830, 0.1018]],</span>
<span class="go"> [[ 1.9348, -2.5744],</span>
<span class="go"> [-2.5744, 4.6386]]], dtype=torch.float64)</span>
<span class="gp">>>> </span><span class="n">l</span>
<span class="go">tensor([[[ 1.0784, 0.0000],</span>
<span class="go"> [ 1.8766, 1.7713]],</span>
<span class="go"> [[ 0.6471, 0.0000],</span>
<span class="go"> [ 0.2829, 0.1477]],</span>
<span class="go"> [[ 1.3910, 0.0000],</span>
<span class="go"> [-1.8509, 1.1014]]], dtype=torch.float64)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">allclose</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">l</span><span class="p">,</span> <span class="n">l</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="o">-</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">)),</span> <span class="n">a</span><span class="p">)</span>
<span class="go">True</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.cond">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">cond</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">p=None</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None</em><span class="sig-paren">)</span> → Tensor<a class="headerlink" href="#torch.linalg.cond" title="Permalink to this definition">¶</a></dt>
<dd><p>Computes the condition number of a matrix <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, or of each matrix in
a batched <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, using the matrix norm defined by <code class="xref py py-attr docutils literal notranslate"><span class="pre">p</span></code>.</p>
<p>For norms <cite>{‘fro’, ‘nuc’, inf, -inf, 1, -1}</cite> this is defined as the matrix norm of <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>
times the matrix norm of the inverse of <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> computed using <a class="reference internal" href="#torch.linalg.norm" title="torch.linalg.norm"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.norm()</span></code></a>. While
for norms <cite>{None, 2, -2}</cite> this is defined as the ratio between the largest and smallest singular
values computed using <a class="reference internal" href="#torch.linalg.svd" title="torch.linalg.svd"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.svd()</span></code></a>.</p>
<p>This function supports float, double, cfloat and cdouble dtypes.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function may synchronize that device with the CPU depending
on which norm <code class="xref py py-attr docutils literal notranslate"><span class="pre">p</span></code> is used.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>For norms <cite>{None, 2, -2}</cite>, <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> may be a non-square matrix or batch of non-square matrices.
For other norms, however, <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> must be a square matrix or a batch of square matrices,
and if this requirement is not satisfied a RuntimeError will be thrown.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>For norms <cite>{‘fro’, ‘nuc’, inf, -inf, 1, -1}</cite> if <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is a non-invertible matrix then
a tensor containing infinity will be returned. If <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is a batch of matrices and one
or more of them is not invertible then a RuntimeError will be thrown.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the input matrix of size <cite>(m, n)</cite> or the batch of matrices of size <cite>(*, m, n)</cite>
where <cite>*</cite> is one or more batch dimensions.</p></li>
<li><p><strong>p</strong> (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.9)"><em>int</em></a><em>, </em><a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.9)"><em>float</em></a><em>, </em><em>inf</em><em>, </em><em>-inf</em><em>, </em><em>'fro'</em><em>, </em><em>'nuc'</em><em>, </em><em>optional</em>) – <p>the type of the matrix norm to use in the computations.
inf refers to <code class="xref py py-attr docutils literal notranslate"><span class="pre">float('inf')</span></code>, numpy’s <code class="xref py py-attr docutils literal notranslate"><span class="pre">inf</span></code> object, or any equivalent object.
The following norms can be used:</p>
<table class="docutils colwidths-auto align-default">
<thead>
<tr class="row-odd"><th class="head"><p>p</p></th>
<th class="head"><p>norm for matrices</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p>None</p></td>
<td><p>ratio of the largest singular value to the smallest singular value</p></td>
</tr>
<tr class="row-odd"><td><p>’fro’</p></td>
<td><p>Frobenius norm</p></td>
</tr>
<tr class="row-even"><td><p>’nuc’</p></td>
<td><p>nuclear norm</p></td>
</tr>
<tr class="row-odd"><td><p>inf</p></td>
<td><p>max(sum(abs(x), dim=1))</p></td>
</tr>
<tr class="row-even"><td><p>-inf</p></td>
<td><p>min(sum(abs(x), dim=1))</p></td>
</tr>
<tr class="row-odd"><td><p>1</p></td>
<td><p>max(sum(abs(x), dim=0))</p></td>
</tr>
<tr class="row-even"><td><p>-1</p></td>
<td><p>min(sum(abs(x), dim=0))</p></td>
</tr>
<tr class="row-odd"><td><p>2</p></td>
<td><p>ratio of the largest singular value to the smallest singular value</p></td>
</tr>
<tr class="row-even"><td><p>-2</p></td>
<td><p>ratio of the smallest singular value to the largest singular value</p></td>
</tr>
</tbody>
</table>
<p>Default: <code class="docutils literal notranslate"><span class="pre">None</span></code></p>
</p></li>
</ul>
</dd>
<dt class="field-even">Keyword Arguments</dt>
<dd class="field-even"><p><strong>out</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a><em>, </em><em>optional</em>) – tensor to write the output to. Default is <code class="docutils literal notranslate"><span class="pre">None</span></code>.</p>
</dd>
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>The condition number of <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>. The output dtype is always real valued
even for complex inputs (e.g. float if <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is cfloat).</p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([[</span><span class="mf">1.</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([1.4142])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="s1">'fro'</span><span class="p">)</span>
<span class="go">tensor(3.1623)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="s1">'nuc'</span><span class="p">)</span>
<span class="go">tensor(9.2426)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="nb">float</span><span class="p">(</span><span class="s1">'inf'</span><span class="p">))</span>
<span class="go">tensor(2.)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="nb">float</span><span class="p">(</span><span class="s1">'-inf'</span><span class="p">))</span>
<span class="go">tensor(1.)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span>
<span class="go">tensor(2.)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span>
<span class="go">tensor(1.)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="go">tensor([1.4142])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="o">-</span><span class="mi">2</span><span class="p">)</span>
<span class="go">tensor([0.7071])</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[[-0.9204, 1.1140, 1.2055],</span>
<span class="go"> [ 0.3988, -0.2395, -0.7441],</span>
<span class="go"> [-0.5160, 0.3115, 0.2619]],</span>
<span class="go"> [[-2.2128, 0.9241, 2.1492],</span>
<span class="go"> [-1.1277, 2.7604, -0.8760],</span>
<span class="go"> [ 1.2159, 0.5960, 0.0498]]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([[9.5917],</span>
<span class="go"> [3.2538]])</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[[-0.4671-0.2137j, -0.1334-0.9508j, 0.6252+0.1759j],</span>
<span class="go"> [-0.3486-0.2991j, -0.1317+0.1252j, 0.3025-0.1604j],</span>
<span class="go"> [-0.5634+0.8582j, 0.1118-0.4677j, -0.1121+0.7574j]],</span>
<span class="go"> [[ 0.3964+0.2533j, 0.9385-0.6417j, -0.0283-0.8673j],</span>
<span class="go"> [ 0.2635+0.2323j, -0.8929-1.1269j, 0.3332+0.0733j],</span>
<span class="go"> [ 0.1151+0.1644j, -1.1163+0.3471j, -0.5870+0.1629j]]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([[4.6245],</span>
<span class="go"> [4.5671]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">cond</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span>
<span class="go">tensor([9.2589, 9.3486])</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.det">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">det</code><span class="sig-paren">(</span><em class="sig-param">input</em><span class="sig-paren">)</span> → Tensor<a class="headerlink" href="#torch.linalg.det" title="Permalink to this definition">¶</a></dt>
<dd><p>Computes the determinant of a square matrix <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, or of each square matrix
in a batched <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>.</p>
<p>This function supports float, double, cfloat and cdouble dtypes.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function synchronizes that device with the CPU.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The determinant is computed using LU factorization. LAPACK’s <cite>getrf</cite> is used for CPU inputs,
and MAGMA’s <cite>getrf</cite> is used for CUDA inputs.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Backward through <cite>det</cite> internally uses <a class="reference internal" href="#torch.linalg.svd" title="torch.linalg.svd"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.svd()</span></code></a> when <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is not
invertible. In this case, double backward through <cite>det</cite> will be unstable when <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>
doesn’t have distinct singular values. See <a class="reference internal" href="#torch.linalg.svd" title="torch.linalg.svd"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.svd()</span></code></a> for more details.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the input matrix of size <cite>(n, n)</cite> or the batch of matrices of size
<cite>(*, n, n)</cite> where <cite>*</cite> is one or more batch dimensions.</p>
</dd>
</dl>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[ 0.9478, 0.9158, -1.1295],</span>
<span class="go"> [ 0.9701, 0.7346, -1.8044],</span>
<span class="go"> [-0.2337, 0.0557, 0.6929]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">det</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor(0.0934)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[[ 0.9254, -0.6213],</span>
<span class="go"> [-0.5787, 1.6843]],</span>
<span class="go"> [[ 0.3242, -0.9665],</span>
<span class="go"> [ 0.4539, -0.0887]],</span>
<span class="go"> [[ 1.1336, -0.4025],</span>
<span class="go"> [-0.7089, 0.9032]]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">det</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([1.1990, 0.4099, 0.7386])</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.slogdet">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">slogdet</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None) -> (Tensor</em>, <em class="sig-param">Tensor</em><span class="sig-paren">)</span><a class="headerlink" href="#torch.linalg.slogdet" title="Permalink to this definition">¶</a></dt>
<dd><p>Calculates the sign and natural logarithm of the absolute value of a square matrix’s determinant,
or of the absolute values of the determinants of a batch of square matrices <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>.
The determinant can be computed with <code class="docutils literal notranslate"><span class="pre">sign</span> <span class="pre">*</span> <span class="pre">exp(logabsdet)</span></code>.</p>
<p>Supports input of float, double, cfloat and cdouble datatypes.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function synchronizes that device with the CPU.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The determinant is computed using LU factorization. LAPACK’s <cite>getrf</cite> is used for CPU inputs,
and MAGMA’s <cite>getrf</cite> is used for CUDA inputs.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>For matrices that have zero determinant, this returns <code class="docutils literal notranslate"><span class="pre">(0,</span> <span class="pre">-inf)</span></code>.
If <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is batched then the entries in the result tensors corresponding to matrices with
the zero determinant have sign 0 and the natural logarithm of the absolute value of the determinant -inf.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the input matrix of size <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mo stretchy="false">(</mo><mi>n</mi><mo separator="true">,</mo><mi>n</mi><mo stretchy="false">)</mo></mrow><annotation encoding="application/x-tex">(n, n)</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:1em;vertical-align:-0.25em;"></span><span class="mopen">(</span><span class="mord mathnormal">n</span><span class="mpunct">,</span><span class="mspace" style="margin-right:0.16666666666666666em;"></span><span class="mord mathnormal">n</span><span class="mclose">)</span></span></span></span>
</span> or the batch of matrices of size <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mo stretchy="false">(</mo><mo>∗</mo><mo separator="true">,</mo><mi>n</mi><mo separator="true">,</mo><mi>n</mi><mo stretchy="false">)</mo></mrow><annotation encoding="application/x-tex">(*, n, n)</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:1em;vertical-align:-0.25em;"></span><span class="mopen">(</span><span class="mord">∗</span><span class="mpunct">,</span><span class="mspace" style="margin-right:0.16666666666666666em;"></span><span class="mord mathnormal">n</span><span class="mpunct">,</span><span class="mspace" style="margin-right:0.16666666666666666em;"></span><span class="mord mathnormal">n</span><span class="mclose">)</span></span></span></span>
</span>
where <span class="math"><span class="katex"><span class="katex-mathml"><math xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mo>∗</mo></mrow><annotation encoding="application/x-tex">*</annotation></semantics></math></span><span class="katex-html" aria-hidden="true"><span class="base"><span class="strut" style="height:0.46528em;vertical-align:0em;"></span><span class="mord">∗</span></span></span></span>
</span> is one or more batch dimensions.</p>
</dd>
<dt class="field-even">Keyword Arguments</dt>
<dd class="field-even"><p><strong>out</strong> (<a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#tuple" title="(in Python v3.9)"><em>tuple</em></a><em>, </em><em>optional</em>) – tuple of two tensors to write the output to.</p>
</dd>
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p>A namedtuple (sign, logabsdet) containing the sign of the determinant and the natural logarithm
of the absolute value of determinant, respectively.</p>
</dd>
</dl>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">A</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">A</span>
<span class="go">tensor([[ 0.0032, -0.2239, -1.1219],</span>
<span class="go"> [-0.6690, 0.1161, 0.4053],</span>
<span class="go"> [-1.6218, -0.9273, -0.0082]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">det</span><span class="p">(</span><span class="n">A</span><span class="p">)</span>
<span class="go">tensor(-0.7576)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">logdet</span><span class="p">(</span><span class="n">A</span><span class="p">)</span>
<span class="go">tensor(nan)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">slogdet</span><span class="p">(</span><span class="n">A</span><span class="p">)</span>
<span class="go">torch.return_types.linalg_slogdet(sign=tensor(-1.), logabsdet=tensor(-0.2776))</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.eigh">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">eigh</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">UPLO='L'</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None) -> (Tensor</em>, <em class="sig-param">Tensor</em><span class="sig-paren">)</span><a class="headerlink" href="#torch.linalg.eigh" title="Permalink to this definition">¶</a></dt>
<dd><p>Computes the eigenvalues and eigenvectors of a complex Hermitian (or real symmetric)
matrix <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, or of each such matrix in a batched <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>.</p>
<p>For a single matrix <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, the tensor of eigenvalues <cite>w</cite> and the tensor of eigenvectors
<cite>V</cite> decompose the <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> such that <cite>input = V diag(w) Vᴴ</cite>, where <cite>Vᴴ</cite> is the transpose of <cite>V</cite>
for real-valued <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, or the conjugate transpose of <cite>V</cite> for complex-valued <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>.</p>
<p>Since the matrix or matrices in <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> are assumed to be Hermitian, the imaginary part of their diagonals
is always treated as zero. When <code class="xref py py-attr docutils literal notranslate"><span class="pre">UPLO</span></code> is “L”, its default value, only the lower triangular part of each
matrix is used in the computation. When <code class="xref py py-attr docutils literal notranslate"><span class="pre">UPLO</span></code> is “U” only the upper triangular part of each matrix is used.</p>
<p>Supports input of float, double, cfloat and cdouble dtypes.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function synchronizes that device with the CPU.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The eigenvalues/eigenvectors are computed using LAPACK’s <cite>syevd</cite> and <cite>heevd</cite> routines for CPU inputs,
and MAGMA’s <cite>syevd</cite> and <cite>heevd</cite> routines for CUDA inputs.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The eigenvalues of real symmetric or complex Hermitian matrices are always real.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The eigenvectors of matrices are not unique, so any eigenvector multiplied by a constant remains
a valid eigenvector. This function may compute different eigenvector representations on
different device types. Usually the difference is only in the sign of the eigenvector.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>See <a class="reference internal" href="#torch.linalg.eigvalsh" title="torch.linalg.eigvalsh"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.eigvalsh()</span></code></a> for a related function that computes only eigenvalues.
However, that function is not differentiable.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the Hermitian <cite>n times n</cite> matrix or the batch of such matrices of size
<cite>(*, n, n)</cite> where <cite>*</cite> is one or more batch dimensions.</p></li>
<li><p><strong>UPLO</strong> (<em>'L'</em><em>, </em><em>'U'</em><em>, </em><em>optional</em>) – controls whether to use the upper-triangular or the lower-triangular part
of <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> in the computations. Default is <code class="docutils literal notranslate"><span class="pre">'L'</span></code>.</p></li>
</ul>
</dd>
<dt class="field-even">Keyword Arguments</dt>
<dd class="field-even"><p><strong>out</strong> (<a class="reference external" href="https://docs.python.org/3/library/stdtypes.html#tuple" title="(in Python v3.9)"><em>tuple</em></a><em>, </em><em>optional</em>) – tuple of two tensors to write the output to. Default is <code class="docutils literal notranslate"><span class="pre">None</span></code>.</p>
</dd>
<dt class="field-odd">Returns</dt>
<dd class="field-odd"><p><p>A namedtuple (eigenvalues, eigenvectors) containing</p>
<blockquote>
<div><ul class="simple">
<li><dl class="simple">
<dt><strong>eigenvalues</strong> (<em>Tensor</em>): Shape <cite>(*, m)</cite>.</dt><dd><p>The eigenvalues in ascending order.</p>
</dd>
</dl>
</li>
<li><dl class="simple">
<dt><strong>eigenvectors</strong> (<em>Tensor</em>): Shape <cite>(*, m, m)</cite>.</dt><dd><p>The orthonormal eigenvectors of the <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>.</p>
</dd>
</dl>
</li>
</ul>
</div></blockquote>
</p>
</dd>
<dt class="field-even">Return type</dt>
<dd class="field-even"><p>(<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor">Tensor</a>, <a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor">Tensor</a>)</p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex128</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">a</span> <span class="o">+</span> <span class="n">a</span><span class="o">.</span><span class="n">t</span><span class="p">()</span><span class="o">.</span><span class="n">conj</span><span class="p">()</span> <span class="c1"># creates a Hermitian matrix</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[2.9228+0.0000j, 0.2029-0.0862j],</span>
<span class="go"> [0.2029+0.0862j, 0.3464+0.0000j]], dtype=torch.complex128)</span>
<span class="gp">>>> </span><span class="n">w</span><span class="p">,</span> <span class="n">v</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">eigh</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">w</span>
<span class="go">tensor([0.3277, 2.9415], dtype=torch.float64)</span>
<span class="gp">>>> </span><span class="n">v</span>
<span class="go">tensor([[-0.0846+-0.0000j, -0.9964+0.0000j],</span>
<span class="go"> [ 0.9170+0.3898j, -0.0779-0.0331j]], dtype=torch.complex128)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">allclose</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">v</span><span class="p">,</span> <span class="n">torch</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">w</span><span class="o">.</span><span class="n">to</span><span class="p">(</span><span class="n">v</span><span class="o">.</span><span class="n">dtype</span><span class="p">)</span><span class="o">.</span><span class="n">diag_embed</span><span class="p">(),</span> <span class="n">v</span><span class="o">.</span><span class="n">t</span><span class="p">()</span><span class="o">.</span><span class="n">conj</span><span class="p">())),</span> <span class="n">a</span><span class="p">)</span>
<span class="go">True</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">float64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">a</span> <span class="o">+</span> <span class="n">a</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="o">-</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span> <span class="c1"># creates a symmetric matrix</span>
<span class="gp">>>> </span><span class="n">w</span><span class="p">,</span> <span class="n">v</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">eigh</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">allclose</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">v</span><span class="p">,</span> <span class="n">torch</span><span class="o">.</span><span class="n">matmul</span><span class="p">(</span><span class="n">w</span><span class="o">.</span><span class="n">diag_embed</span><span class="p">(),</span> <span class="n">v</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="o">-</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">))),</span> <span class="n">a</span><span class="p">)</span>
<span class="go">True</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.eigvalsh">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">eigvalsh</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">UPLO='L'</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None</em><span class="sig-paren">)</span> → Tensor<a class="headerlink" href="#torch.linalg.eigvalsh" title="Permalink to this definition">¶</a></dt>
<dd><p>Computes the eigenvalues of a complex Hermitian (or real symmetric) matrix <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>,
or of each such matrix in a batched <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>. The eigenvalues are returned in ascending order.</p>
<p>Since the matrix or matrices in <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> are assumed to be Hermitian, the imaginary part of their diagonals
is always treated as zero. When <code class="xref py py-attr docutils literal notranslate"><span class="pre">UPLO</span></code> is “L”, its default value, only the lower triangular part of
each matrix is used in the computation. When <code class="xref py py-attr docutils literal notranslate"><span class="pre">UPLO</span></code> is “U” only the upper triangular part of each matrix is used.</p>
<p>Supports input of float, double, cfloat and cdouble dtypes.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function synchronizes that device with the CPU.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The eigenvalues are computed using LAPACK’s <cite>syevd</cite> and <cite>heevd</cite> routines for CPU inputs,
and MAGMA’s <cite>syevd</cite> and <cite>heevd</cite> routines for CUDA inputs.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The eigenvalues of real symmetric or complex Hermitian matrices are always real.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>This function doesn’t support backpropagation, please use <a class="reference internal" href="#torch.linalg.eigh" title="torch.linalg.eigh"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.eigh()</span></code></a> instead,
which also computes the eigenvectors.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>See <a class="reference internal" href="#torch.linalg.eigh" title="torch.linalg.eigh"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.eigh()</span></code></a> for a related function that computes both eigenvalues and eigenvectors.</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the Hermitian <cite>n times n</cite> matrix or the batch
of such matrices of size <cite>(*, n, n)</cite> where <cite>*</cite> is one or more batch dimensions.</p></li>
<li><p><strong>UPLO</strong> (<em>'L'</em><em>, </em><em>'U'</em><em>, </em><em>optional</em>) – controls whether to use the upper-triangular or the lower-triangular part
of <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> in the computations. Default is <code class="docutils literal notranslate"><span class="pre">'L'</span></code>.</p></li>
</ul>
</dd>
<dt class="field-even">Keyword Arguments</dt>
<dd class="field-even"><p><strong>out</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a><em>, </em><em>optional</em>) – tensor to write the output to. Default is <code class="docutils literal notranslate"><span class="pre">None</span></code>.</p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex128</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">a</span> <span class="o">+</span> <span class="n">a</span><span class="o">.</span><span class="n">t</span><span class="p">()</span><span class="o">.</span><span class="n">conj</span><span class="p">()</span> <span class="c1"># creates a Hermitian matrix</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[2.9228+0.0000j, 0.2029-0.0862j],</span>
<span class="go"> [0.2029+0.0862j, 0.3464+0.0000j]], dtype=torch.complex128)</span>
<span class="gp">>>> </span><span class="n">w</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">eigvalsh</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">w</span>
<span class="go">tensor([0.3277, 2.9415], dtype=torch.float64)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">float64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">a</span> <span class="o">+</span> <span class="n">a</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="o">-</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span> <span class="c1"># creates a symmetric matrix</span>
<span class="gp">>>> </span><span class="n">a</span>
<span class="go">tensor([[[ 2.8050, -0.3850],</span>
<span class="go"> [-0.3850, 3.2376]],</span>
<span class="go"> [[-1.0307, -2.7457],</span>
<span class="go"> [-2.7457, -1.7517]],</span>
<span class="go"> [[ 1.7166, 2.2207],</span>
<span class="go"> [ 2.2207, -2.0898]]], dtype=torch.float64)</span>
<span class="gp">>>> </span><span class="n">w</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">eigvalsh</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">w</span>
<span class="go">tensor([[ 2.5797, 3.4629],</span>
<span class="go"> [-4.1605, 1.3780],</span>
<span class="go"> [-3.1113, 2.7381]], dtype=torch.float64)</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.matrix_rank">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">matrix_rank</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">tol=None</em>, <em class="sig-param">hermitian=False</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None</em><span class="sig-paren">)</span> → Tensor<a class="headerlink" href="#torch.linalg.matrix_rank" title="Permalink to this definition">¶</a></dt>
<dd><p>Computes the numerical rank of a matrix <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>, or of each matrix in a batched <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>.</p>
<p>The matrix rank is computed as the number of singular values (or absolute eigenvalues when <code class="xref py py-attr docutils literal notranslate"><span class="pre">hermitian</span></code> is <code class="docutils literal notranslate"><span class="pre">True</span></code>)
that are greater than the specified <code class="xref py py-attr docutils literal notranslate"><span class="pre">tol</span></code> threshold.</p>
<p>If <code class="xref py py-attr docutils literal notranslate"><span class="pre">tol</span></code> is not specified, <code class="xref py py-attr docutils literal notranslate"><span class="pre">tol</span></code> is set to <code class="docutils literal notranslate"><span class="pre">S.max(dim=-1)*max(input.shape[-2:])*eps</span></code>,
where <code class="docutils literal notranslate"><span class="pre">S</span></code> is the singular values (or absolute eigenvalues when <code class="xref py py-attr docutils literal notranslate"><span class="pre">hermitian</span></code> is <code class="docutils literal notranslate"><span class="pre">True</span></code>), and
<code class="docutils literal notranslate"><span class="pre">eps</span></code> is the epsilon value for the datatype of <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code>. The epsilon value can be obtained using
the <code class="docutils literal notranslate"><span class="pre">eps</span></code> attribute of <code class="xref py py-class docutils literal notranslate"><span class="pre">torch.finfo</span></code>.</p>
<p>Supports input of float, double, cfloat and cdouble dtypes.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>When given inputs on a CUDA device, this function synchronizes that device with the CPU.</p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The matrix rank is computed using singular value decomposition (see <a class="reference internal" href="#torch.linalg.svd" title="torch.linalg.svd"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.svd()</span></code></a>) by default.
If <code class="xref py py-attr docutils literal notranslate"><span class="pre">hermitian</span></code> is <code class="docutils literal notranslate"><span class="pre">True</span></code>, then <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is assumed to be Hermitian (symmetric if real-valued),
and the computation is done by obtaining the eigenvalues (see <a class="reference internal" href="#torch.linalg.eigvalsh" title="torch.linalg.eigvalsh"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.linalg.eigvalsh()</span></code></a>).</p>
</div>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – the input matrix of size <cite>(m, n)</cite> or the batch of matrices of size <cite>(*, m, n)</cite>
where <cite>*</cite> is one or more batch dimensions.</p></li>
<li><p><strong>tol</strong> (<a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.9)"><em>float</em></a><em>, </em><em>optional</em>) – the tolerance value. Default is <code class="docutils literal notranslate"><span class="pre">None</span></code></p></li>
<li><p><strong>hermitian</strong> (<a class="reference external" href="https://docs.python.org/3/library/functions.html#bool" title="(in Python v3.9)"><em>bool</em></a><em>, </em><em>optional</em>) – indicates whether <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is Hermitian. Default is <code class="docutils literal notranslate"><span class="pre">False</span></code>.</p></li>
</ul>
</dd>
<dt class="field-even">Keyword Arguments</dt>
<dd class="field-even"><p><strong>out</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a><em>, </em><em>optional</em>) – tensor to write the output to. Default is <code class="docutils literal notranslate"><span class="pre">None</span></code>.</p>
</dd>
</dl>
<p>Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">eye</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor(10)</span>
<span class="gp">>>> </span><span class="n">b</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">eye</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">b</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">0</span><span class="p">]</span> <span class="o">=</span> <span class="mi">0</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">b</span><span class="p">)</span>
<span class="go">tensor(9)</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">4</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([2, 2, 2, 2])</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([[2, 2, 2, 2],</span>
<span class="go"> [2, 2, 2, 2]])</span>
<span class="gp">>>> </span><span class="n">a</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">)</span>
<span class="go">tensor([[3, 3, 3, 3],</span>
<span class="go"> [3, 3, 3, 3]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">hermitian</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="go">tensor([[3, 3, 3, 3],</span>
<span class="go"> [3, 3, 3, 3]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">tol</span><span class="o">=</span><span class="mf">1.0</span><span class="p">)</span>
<span class="go">tensor([[3, 2, 2, 2],</span>
<span class="go"> [1, 2, 1, 2]])</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">linalg</span><span class="o">.</span><span class="n">matrix_rank</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">tol</span><span class="o">=</span><span class="mf">1.0</span><span class="p">,</span> <span class="n">hermitian</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="go">tensor([[2, 2, 2, 1],</span>
<span class="go"> [1, 2, 2, 2]])</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.linalg.norm">
<code class="sig-prename descclassname">torch.linalg.</code><code class="sig-name descname">norm</code><span class="sig-paren">(</span><em class="sig-param">input</em>, <em class="sig-param">ord=None</em>, <em class="sig-param">dim=None</em>, <em class="sig-param">keepdim=False</em>, <em class="sig-param">*</em>, <em class="sig-param">out=None</em>, <em class="sig-param">dtype=None</em><span class="sig-paren">)</span> → Tensor<a class="headerlink" href="#torch.linalg.norm" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns the matrix norm or vector norm of a given tensor.</p>
<p>This function can calculate one of eight different types of matrix norms, or one
of an infinite number of vector norms, depending on both the number of reduction
dimensions and the value of the <cite>ord</cite> parameter.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>input</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – The input tensor. If dim is None, x must be 1-D or 2-D, unless <code class="xref py py-attr docutils literal notranslate"><span class="pre">ord</span></code>
is None. If both <code class="xref py py-attr docutils literal notranslate"><span class="pre">dim</span></code> and <code class="xref py py-attr docutils literal notranslate"><span class="pre">ord</span></code> are None, the 2-norm of the input flattened to 1-D
will be returned. Its data type must be either a floating point or complex type. For complex
inputs, the norm is calculated on of the absolute values of each element. If the input is
complex and neither <code class="xref py py-attr docutils literal notranslate"><span class="pre">dtype</span></code> nor <code class="xref py py-attr docutils literal notranslate"><span class="pre">out</span></code> is specified, the result’s data type will
be the corresponding floating point type (e.g. float if <code class="xref py py-attr docutils literal notranslate"><span class="pre">input</span></code> is complexfloat).</p></li>
<li><p><strong>ord</strong> (<a class="reference external" href="https://docs.python.org/3/library/functions.html#int" title="(in Python v3.9)"><em>int</em></a><em>, </em><a class="reference external" href="https://docs.python.org/3/library/functions.html#float" title="(in Python v3.9)"><em>float</em></a><em>, </em><em>inf</em><em>, </em><em>-inf</em><em>, </em><em>'fro'</em><em>, </em><em>'nuc'</em><em>, </em><em>optional</em>) – <p>The order of norm.
inf refers to <code class="xref py py-attr docutils literal notranslate"><span class="pre">float('inf')</span></code>, numpy’s <code class="xref py py-attr docutils literal notranslate"><span class="pre">inf</span></code> object, or any equivalent object.
The following norms can be calculated:</p>
<table class="docutils colwidths-auto align-default">
<thead>
<tr class="row-odd"><th class="head"><p>ord</p></th>
<th class="head"><p>norm for matrices</p></th>
<th class="head"><p>norm for vectors</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p>None</p></td>
<td><p>Frobenius norm</p></td>
<td><p>2-norm</p></td>
</tr>
<tr class="row-odd"><td><p>’fro’</p></td>
<td><p>Frobenius norm</p></td>
<td><p>– not supported –</p></td>
</tr>
<tr class="row-even"><td><p>‘nuc’</p></td>
<td><p>nuclear norm</p></td>
<td><p>– not supported –</p></td>
</tr>
<tr class="row-odd"><td><p>inf</p></td>
<td><p>max(sum(abs(x), dim=1))</p></td>
<td><p>max(abs(x))</p></td>
</tr>
<tr class="row-even"><td><p>-inf</p></td>
<td><p>min(sum(abs(x), dim=1))</p></td>