forked from scikit-learn/scikit-learn.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathadvanced_installation.html
698 lines (663 loc) · 45.5 KB
/
advanced_installation.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
<!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 charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Installing the development version of scikit-learn — scikit-learn 1.0.dev0 documentation</title>
<link rel="canonical" href="http://scikit-learn.org/stable/developers/advanced_installation.html" />
<link rel="shortcut icon" href="../_static/favicon.ico"/>
<link rel="stylesheet" href="../_static/css/vendor/bootstrap.min.css" type="text/css" />
<link rel="stylesheet" href="../_static/gallery.css" type="text/css" />
<link rel="stylesheet" href="../_static/gallery-binder.css" type="text/css" />
<link rel="stylesheet" href="../_static/gallery-dataframe.css" type="text/css" />
<link rel="stylesheet" href="../_static/gallery-rendered-html.css" type="text/css" />
<link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
<script id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script>
<script src="../_static/jquery.js"></script>
</head>
<body>
<nav id="navbar" class="sk-docs-navbar navbar navbar-expand-md navbar-light bg-light py-0">
<div class="container-fluid sk-docs-container px-0">
<a class="navbar-brand py-0" href="../index.html">
<img
class="sk-brand-img"
src="../_static/scikit-learn-logo-small.png"
alt="logo"/>
</a>
<button
id="sk-navbar-toggler"
class="navbar-toggler"
type="button"
data-toggle="collapse"
data-target="#navbarSupportedContent"
aria-controls="navbarSupportedContent"
aria-expanded="false"
aria-label="Toggle navigation"
>
<span class="navbar-toggler-icon"></span>
</button>
<div class="sk-navbar-collapse collapse navbar-collapse" id="navbarSupportedContent">
<ul class="navbar-nav mr-auto">
<li class="nav-item">
<a class="sk-nav-link nav-link" href="../install.html">Install</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link" href="../user_guide.html">User Guide</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link" href="../modules/classes.html">API</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link" href="../auto_examples/index.html">Examples</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../getting_started.html">Getting Started</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../tutorial/index.html">Tutorial</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../whats_new/v1.0.html">What's new</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../glossary.html">Glossary</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="index.html">Development</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../faq.html">FAQ</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../support.html">Support</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../related_projects.html">Related packages</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../roadmap.html">Roadmap</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="../about.html">About us</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="https://github.com/scikit-learn/scikit-learn">GitHub</a>
</li>
<li class="nav-item">
<a class="sk-nav-link nav-link nav-more-item-mobile-items" href="https://scikit-learn.org/dev/versions.html">Other Versions and Download</a>
</li>
<li class="nav-item dropdown nav-more-item-dropdown">
<a class="sk-nav-link nav-link dropdown-toggle" href="#" id="navbarDropdown" role="button" data-toggle="dropdown" aria-haspopup="true" aria-expanded="false">More</a>
<div class="dropdown-menu" aria-labelledby="navbarDropdown">
<a class="sk-nav-dropdown-item dropdown-item" href="../getting_started.html">Getting Started</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../tutorial/index.html">Tutorial</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../whats_new/v1.0.html">What's new</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../glossary.html">Glossary</a>
<a class="sk-nav-dropdown-item dropdown-item" href="index.html">Development</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../faq.html">FAQ</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../support.html">Support</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../related_projects.html">Related packages</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../roadmap.html">Roadmap</a>
<a class="sk-nav-dropdown-item dropdown-item" href="../about.html">About us</a>
<a class="sk-nav-dropdown-item dropdown-item" href="https://github.com/scikit-learn/scikit-learn">GitHub</a>
<a class="sk-nav-dropdown-item dropdown-item" href="https://scikit-learn.org/dev/versions.html">Other Versions and Download</a>
</div>
</li>
</ul>
<div id="searchbox" role="search">
<div class="searchformwrapper">
<form class="search" action="../search.html" method="get">
<input class="sk-search-text-input" type="text" name="q" aria-labelledby="searchlabel" />
<input class="sk-search-text-btn" type="submit" value="Go" />
</form>
</div>
</div>
</div>
</div>
</nav>
<div class="d-flex" id="sk-doc-wrapper">
<input type="checkbox" name="sk-toggle-checkbox" id="sk-toggle-checkbox">
<label id="sk-sidemenu-toggle" class="sk-btn-toggle-toc btn sk-btn-primary" for="sk-toggle-checkbox">Toggle Menu</label>
<div id="sk-sidebar-wrapper" class="border-right">
<div class="sk-sidebar-toc-wrapper">
<div class="sk-sidebar-toc-logo">
<a href="../index.html">
<img
class="sk-brand-img"
src="../_static/scikit-learn-logo-small.png"
alt="logo"/>
</a>
</div>
<div class="btn-group w-100 mb-2" role="group" aria-label="rellinks">
<a href="performance.html" role="button" class="btn sk-btn-rellink py-1" sk-rellink-tooltip="How to optimize for speed">Prev</a><a href="index.html" role="button" class="btn sk-btn-rellink py-1" sk-rellink-tooltip="Developer’s Guide">Up</a>
<a href="bug_triaging.html" role="button" class="btn sk-btn-rellink py-1" sk-rellink-tooltip="Bug triaging and issue curation">Next</a>
</div>
<div class="alert alert-danger p-1 mb-2" role="alert">
<p class="text-center mb-0">
<strong>scikit-learn 1.0.dev0</strong><br/>
<a href="http://scikit-learn.org/dev/versions.html">Other versions</a>
</p>
</div>
<div class="alert alert-warning p-1 mb-2" role="alert">
<p class="text-center mb-0">
Please <a class="font-weight-bold" href="../about.html#citing-scikit-learn"><string>cite us</string></a> if you use the software.
</p>
</div>
<div class="sk-sidebar-toc">
<ul>
<li><a class="reference internal" href="#">Installing the development version of scikit-learn</a><ul>
<li><a class="reference internal" href="#installing-nightly-builds">Installing nightly builds</a></li>
<li><a class="reference internal" href="#building-from-source">Building from source</a><ul>
<li><a class="reference internal" href="#dependencies">Dependencies</a><ul>
<li><a class="reference internal" href="#runtime-dependencies">Runtime dependencies</a></li>
<li><a class="reference internal" href="#build-dependencies">Build dependencies</a></li>
<li><a class="reference internal" href="#test-dependencies">Test dependencies</a></li>
</ul>
</li>
<li><a class="reference internal" href="#building-a-specific-version-from-a-tag">Building a specific version from a tag</a></li>
<li><a class="reference internal" href="#editable-mode">Editable mode</a></li>
</ul>
</li>
<li><a class="reference internal" href="#platform-specific-instructions">Platform-specific instructions</a><ul>
<li><a class="reference internal" href="#windows">Windows</a></li>
<li><a class="reference internal" href="#macos">macOS</a><ul>
<li><a class="reference internal" href="#macos-compilers-from-conda-forge">macOS compilers from conda-forge</a></li>
<li><a class="reference internal" href="#macos-compilers-from-homebrew">macOS compilers from Homebrew</a></li>
</ul>
</li>
<li><a class="reference internal" href="#linux">Linux</a><ul>
<li><a class="reference internal" href="#linux-compilers-from-the-system">Linux compilers from the system</a></li>
<li><a class="reference internal" href="#linux-compilers-from-conda-forge">Linux compilers from conda-forge</a></li>
</ul>
</li>
<li><a class="reference internal" href="#freebsd">FreeBSD</a></li>
</ul>
</li>
<li><a class="reference internal" href="#alternative-compilers">Alternative compilers</a><ul>
<li><a class="reference internal" href="#building-with-intel-c-compiler-icc-using-oneapi-on-linux">Building with Intel C Compiler (ICC) using oneAPI on Linux</a></li>
</ul>
</li>
<li><a class="reference internal" href="#parallel-builds">Parallel builds</a></li>
</ul>
</li>
</ul>
</div>
</div>
</div>
<div id="sk-page-content-wrapper">
<div class="sk-page-content container-fluid body px-md-3" role="main">
<div class="section" id="installing-the-development-version-of-scikit-learn">
<span id="advanced-installation"></span><h1>Installing the development version of scikit-learn<a class="headerlink" href="#installing-the-development-version-of-scikit-learn" title="Permalink to this headline">¶</a></h1>
<p>This section introduces how to install the <strong>main branch</strong> of scikit-learn.
This can be done by either installing a nightly build or building from source.</p>
<div class="section" id="installing-nightly-builds">
<span id="install-nightly-builds"></span><h2>Installing nightly builds<a class="headerlink" href="#installing-nightly-builds" title="Permalink to this headline">¶</a></h2>
<p>The continuous integration servers of the scikit-learn project build, test
and upload wheel packages for the most recent Python version on a nightly
basis.</p>
<p>Installing a nightly build is the quickest way to:</p>
<ul class="simple">
<li><p>try a new feature that will be shipped in the next release (that is, a
feature from a pull-request that was recently merged to the main branch);</p></li>
<li><p>check whether a bug you encountered has been fixed since the last release.</p></li>
</ul>
<div class="highlight-default notranslate"><div class="highlight"><pre><style type="text/css">
span.prompt1:before {
content: "$ ";
}
</style><span class="prompt1">pip install --pre --extra-index https://pypi.anaconda.org/scipy-wheels-nightly/simple scikit-learn</span>
</pre></div></div></div>
<div class="section" id="building-from-source">
<span id="install-bleeding-edge"></span><h2>Building from source<a class="headerlink" href="#building-from-source" title="Permalink to this headline">¶</a></h2>
<p>Building from source is required to work on a contribution (bug fix, new
feature, code or documentation improvement).</p>
<ol class="arabic" id="git-repo">
<li><p>Use <a class="reference external" href="https://git-scm.com/">Git</a> to check out the latest source from the
<a class="reference external" href="https://github.com/scikit-learn/scikit-learn">scikit-learn repository</a> on
Github.:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">git clone git://github.com/scikit-learn/scikit-learn.git <span class="c1"># add --depth 1 if your connection is slow</span></span>
<span class="prompt1"><span class="nb">cd</span> scikit-learn</span>
</pre></div></div><p>If you plan on submitting a pull-request, you should clone from your fork
instead.</p>
</li>
<li><p>Install a compiler with <a class="reference external" href="https://en.wikipedia.org/wiki/OpenMP">OpenMP</a> support for your platform. See instructions
for <a class="reference internal" href="#compiler-windows"><span class="std std-ref">Windows</span></a>, <a class="reference internal" href="#compiler-macos"><span class="std std-ref">macOS</span></a>, <a class="reference internal" href="#compiler-linux"><span class="std std-ref">Linux</span></a>
and <a class="reference internal" href="#compiler-freebsd"><span class="std std-ref">FreeBSD</span></a>.</p></li>
<li><p>Optional (but recommended): create and activate a dedicated <a class="reference external" href="https://docs.python.org/3/tutorial/venv.html">virtualenv</a>
or <a class="reference external" href="https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html">conda environment</a>.</p></li>
<li><p>Install <a class="reference external" href="https://cython.org">Cython</a> and build the project with pip in <a class="reference internal" href="#editable-mode"><span class="std std-ref">Editable mode</span></a>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">pip install cython</span>
<span class="prompt1">pip install --verbose --no-build-isolation --editable .</span>
</pre></div></div></li>
<li><p>Check that the installed scikit-learn has a version number ending with
<code class="docutils literal notranslate"><span class="pre">.dev0</span></code>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">python -c <span class="s2">"import sklearn; sklearn.show_versions()"</span></span>
</pre></div></div></li>
<li><p>Please refer to the <a class="reference internal" href="index.html#developers-guide"><span class="std std-ref">Developer’s Guide</span></a> and <a class="reference internal" href="tips.html#pytest-tips"><span class="std std-ref">Useful pytest aliases and flags</span></a> to run
the tests on the module of your choice.</p></li>
</ol>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>You will have to run the <code class="docutils literal notranslate"><span class="pre">pip</span> <span class="pre">install</span> <span class="pre">--no-build-isolation</span> <span class="pre">--editable</span> <span class="pre">.</span></code>
command every time the source code of a Cython file is updated
(ending in <code class="docutils literal notranslate"><span class="pre">.pyx</span></code> or <code class="docutils literal notranslate"><span class="pre">.pxd</span></code>). Use the <code class="docutils literal notranslate"><span class="pre">--no-build-isolation</span></code> flag to
avoid compiling the whole project each time, only the files you have
modified.</p>
</div>
<div class="section" id="dependencies">
<h3>Dependencies<a class="headerlink" href="#dependencies" title="Permalink to this headline">¶</a></h3>
<div class="section" id="runtime-dependencies">
<h4>Runtime dependencies<a class="headerlink" href="#runtime-dependencies" title="Permalink to this headline">¶</a></h4>
<p>Scikit-learn requires the following dependencies both at build time and at
runtime:</p>
<ul class="simple">
<li><p>Python (>= 3.6),</p></li>
<li><p>NumPy (>= 1.13.3),</p></li>
<li><p>SciPy (>= 0.19.1),</p></li>
<li><p>Joblib (>= 0.11),</p></li>
<li><p>threadpoolctl (>= 2.0.0).</p></li>
</ul>
<p>Those dependencies are <strong>automatically installed by pip</strong> if they were missing
when building scikit-learn from source.</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>For running on PyPy, PyPy3-v5.10+, Numpy 1.14.0+, and scipy 1.1.0+
are required. For PyPy, only installation instructions with pip apply.</p>
</div>
</div>
<div class="section" id="build-dependencies">
<h4>Build dependencies<a class="headerlink" href="#build-dependencies" title="Permalink to this headline">¶</a></h4>
<p>Building Scikit-learn also requires:</p>
<ul class="simple">
<li><p>Cython >= 0.28.5</p></li>
<li><p>A C/C++ compiler and a matching <a class="reference external" href="https://en.wikipedia.org/wiki/OpenMP">OpenMP</a> runtime library. See the
<a class="reference internal" href="#platform-specific-instructions"><span class="std std-ref">platform system specific instructions</span></a> for more details.</p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>If OpenMP is not supported by the compiler, the build will be done with
OpenMP functionalities disabled. This is not recommended since it will force
some estimators to run in sequential mode instead of leveraging thread-based
parallelism. Setting the <code class="docutils literal notranslate"><span class="pre">SKLEARN_FAIL_NO_OPENMP</span></code> environment variable
(before cythonization) will force the build to fail if OpenMP is not
supported.</p>
</div>
<p>Since version 0.21, scikit-learn automatically detects and use the linear
algebrea library used by SciPy <strong>at runtime</strong>. Scikit-learn has therefore no
build dependency on BLAS/LAPACK implementations such as OpenBlas, Atlas, Blis
or MKL.</p>
</div>
<div class="section" id="test-dependencies">
<h4>Test dependencies<a class="headerlink" href="#test-dependencies" title="Permalink to this headline">¶</a></h4>
<p>Running tests requires:</p>
<ul class="simple">
<li><p>pytest >= 5.0.1</p></li>
</ul>
<p>Some tests also require <a class="reference external" href="https://pandas.pydata.org">pandas</a>.</p>
</div>
</div>
<div class="section" id="building-a-specific-version-from-a-tag">
<h3>Building a specific version from a tag<a class="headerlink" href="#building-a-specific-version-from-a-tag" title="Permalink to this headline">¶</a></h3>
<p>If you want to build a stable version, you can <code class="docutils literal notranslate"><span class="pre">git</span> <span class="pre">checkout</span> <span class="pre"><VERSION></span></code>
to get the code for that particular version, or download an zip archive of
the version from github.</p>
</div>
<div class="section" id="editable-mode">
<span id="id1"></span><h3>Editable mode<a class="headerlink" href="#editable-mode" title="Permalink to this headline">¶</a></h3>
<p>If you run the development version, it is cumbersome to reinstall the package
each time you update the sources. Therefore it is recommended that you install
in with the <code class="docutils literal notranslate"><span class="pre">pip</span> <span class="pre">install</span> <span class="pre">--no-build-isolation</span> <span class="pre">--editable</span> <span class="pre">.</span></code> command, which
allows you to edit the code in-place. This builds the extension in place and
creates a link to the development directory (see <a class="reference external" href="https://pip.pypa.io/en/stable/reference/pip_install/#editable-installs">the pip docs</a>).</p>
<p>This is fundamentally similar to using the command <code class="docutils literal notranslate"><span class="pre">python</span> <span class="pre">setup.py</span> <span class="pre">develop</span></code>
(see <a class="reference external" href="https://setuptools.readthedocs.io/en/latest/setuptools.html#development-mode">the setuptool docs</a>).
It is however preferred to use pip.</p>
<p>On Unix-like systems, you can equivalently type <code class="docutils literal notranslate"><span class="pre">make</span> <span class="pre">in</span></code> from the top-level
folder. Have a look at the <code class="docutils literal notranslate"><span class="pre">Makefile</span></code> for additional utilities.</p>
</div>
</div>
<div class="section" id="platform-specific-instructions">
<span id="id2"></span><h2>Platform-specific instructions<a class="headerlink" href="#platform-specific-instructions" title="Permalink to this headline">¶</a></h2>
<p>Here are instructions to install a working C/C++ compiler with OpenMP support
to build scikit-learn Cython extensions for each supported platform.</p>
<div class="section" id="windows">
<span id="compiler-windows"></span><h3>Windows<a class="headerlink" href="#windows" title="Permalink to this headline">¶</a></h3>
<p>First, install <a class="reference external" href="https://visualstudio.microsoft.com/downloads/">Build Tools for Visual Studio 2019</a>.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>You DO NOT need to install Visual Studio 2019. You only need the “Build
Tools for Visual Studio 2019”, under “All downloads” -> “Tools for Visual
Studio 2019”.</p>
</div>
<p>Secondly, find out if you are running 64-bit or 32-bit Python. The building
command depends on the architecture of the Python interpreter. You can check
the architecture by running the following in <code class="docutils literal notranslate"><span class="pre">cmd</span></code> or <code class="docutils literal notranslate"><span class="pre">powershell</span></code>
console:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">python -c <span class="s2">"import struct; print(struct.calcsize('P') * 8)"</span></span>
</pre></div></div><p>For 64-bit Python, configure the build environment by running the following
commands in <code class="docutils literal notranslate"><span class="pre">cmd</span></code> or an Anaconda Prompt (if you use Anaconda):</p>
<blockquote>
<div><div class="highlight-default notranslate"><div class="highlight"><pre><span></span>$ SET DISTUTILS_USE_SDK=1
$ "C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Auxiliary\Build\vcvarsall.bat" x64
</pre></div>
</div>
</div></blockquote>
<p>Replace <code class="docutils literal notranslate"><span class="pre">x64</span></code> by <code class="docutils literal notranslate"><span class="pre">x86</span></code> to build for 32-bit Python.</p>
<p>Please be aware that the path above might be different from user to user. The
aim is to point to the “vcvarsall.bat” file that will set the necessary
environment variables in the current command prompt.</p>
<p>Finally, build scikit-learn from this command prompt:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">pip install --verbose --no-build-isolation --editable .</span>
</pre></div></div></div>
<div class="section" id="macos">
<span id="compiler-macos"></span><h3>macOS<a class="headerlink" href="#macos" title="Permalink to this headline">¶</a></h3>
<p>The default C compiler on macOS, Apple clang (confusingly aliased as
<code class="docutils literal notranslate"><span class="pre">/usr/bin/gcc</span></code>), does not directly support OpenMP. We present two alternatives
to enable OpenMP support:</p>
<ul class="simple">
<li><p>either install <code class="docutils literal notranslate"><span class="pre">conda-forge::compilers</span></code> with conda;</p></li>
<li><p>or install <code class="docutils literal notranslate"><span class="pre">libomp</span></code> with Homebrew to extend the default Apple clang compiler.</p></li>
</ul>
<p>For Apple Silicon M1 hardware, only the conda-forge method below is known to
work at the time of writing (January 2021). You can install the <code class="docutils literal notranslate"><span class="pre">macos/arm64</span></code>
distribution of conda using the <a class="reference external" href="https://github.com/conda-forge/miniforge#miniforge">miniforge installer</a></p>
<div class="section" id="macos-compilers-from-conda-forge">
<h4>macOS compilers from conda-forge<a class="headerlink" href="#macos-compilers-from-conda-forge" title="Permalink to this headline">¶</a></h4>
<p>If you use the conda package manager (version >= 4.7), you can install the
<code class="docutils literal notranslate"><span class="pre">compilers</span></code> meta-package from the conda-forge channel, which provides
OpenMP-enabled C/C++ compilers based on the llvm toolchain.</p>
<p>First install the macOS command line tools:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">xcode-select --install</span>
</pre></div></div><p>It is recommended to use a dedicated <a class="reference external" href="https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html">conda environment</a> to build
scikit-learn from source:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">conda create -n sklearn-dev -c conda-forge python numpy scipy cython <span class="se">\</span>
joblib threadpoolctl pytest compilers llvm-openmp</span>
<span class="prompt1">conda activate sklearn-dev</span>
<span class="prompt1">make clean</span>
<span class="prompt1">pip install --verbose --no-build-isolation --editable .</span>
</pre></div></div><div class="admonition note">
<p class="admonition-title">Note</p>
<p>If you get any conflicting dependency error message, try commenting out
any custom conda configuration in the <code class="docutils literal notranslate"><span class="pre">$HOME/.condarc</span></code> file. In
particular the <code class="docutils literal notranslate"><span class="pre">channel_priority:</span> <span class="pre">strict</span></code> directive is known to cause
problems for this setup.</p>
</div>
<p>You can check that the custom compilers are properly installed from conda
forge using the following command:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">conda list</span>
</pre></div></div><p>which should include <code class="docutils literal notranslate"><span class="pre">compilers</span></code> and <code class="docutils literal notranslate"><span class="pre">llvm-openmp</span></code>.</p>
<p>The compilers meta-package will automatically set custom environment
variables:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1"><span class="nb">echo</span> <span class="nv">$CC</span></span>
<span class="prompt1"><span class="nb">echo</span> <span class="nv">$CXX</span></span>
<span class="prompt1"><span class="nb">echo</span> <span class="nv">$CFLAGS</span></span>
<span class="prompt1"><span class="nb">echo</span> <span class="nv">$CXXFLAGS</span></span>
<span class="prompt1"><span class="nb">echo</span> <span class="nv">$LDFLAGS</span></span>
</pre></div></div><p>They point to files and folders from your <code class="docutils literal notranslate"><span class="pre">sklearn-dev</span></code> conda environment
(in particular in the bin/, include/ and lib/ subfolders). For instance
<code class="docutils literal notranslate"><span class="pre">-L/path/to/conda/envs/sklearn-dev/lib</span></code> should appear in <code class="docutils literal notranslate"><span class="pre">LDFLAGS</span></code>.</p>
<p>In the log, you should see the compiled extension being built with the clang
and clang++ compilers installed by conda with the <code class="docutils literal notranslate"><span class="pre">-fopenmp</span></code> command line
flag.</p>
</div>
<div class="section" id="macos-compilers-from-homebrew">
<h4>macOS compilers from Homebrew<a class="headerlink" href="#macos-compilers-from-homebrew" title="Permalink to this headline">¶</a></h4>
<p>Another solution is to enable OpenMP support for the clang compiler shipped
by default on macOS.</p>
<p>First install the macOS command line tools:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">xcode-select --install</span>
</pre></div></div><p>Install the <a class="reference external" href="https://brew.sh">Homebrew</a> package manager for macOS.</p>
<p>Install the LLVM OpenMP library:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">brew install libomp</span>
</pre></div></div><p>Set the following environment variables:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1"><span class="nb">export</span> <span class="nv">CC</span><span class="o">=</span>/usr/bin/clang</span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">CXX</span><span class="o">=</span>/usr/bin/clang++</span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">CPPFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$CPPFLAGS</span><span class="s2"> -Xpreprocessor -fopenmp"</span></span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">CFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$CFLAGS</span><span class="s2"> -I/usr/local/opt/libomp/include"</span></span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">CXXFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$CXXFLAGS</span><span class="s2"> -I/usr/local/opt/libomp/include"</span></span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">LDFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$LDFLAGS</span><span class="s2"> -Wl,-rpath,/usr/local/opt/libomp/lib -L/usr/local/opt/libomp/lib -lomp"</span></span>
</pre></div></div><p>Finally, build scikit-learn in verbose mode (to check for the presence of the
<code class="docutils literal notranslate"><span class="pre">-fopenmp</span></code> flag in the compiler commands):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">make clean</span>
<span class="prompt1">pip install --verbose --no-build-isolation --editable .</span>
</pre></div></div></div>
</div>
<div class="section" id="linux">
<span id="compiler-linux"></span><h3>Linux<a class="headerlink" href="#linux" title="Permalink to this headline">¶</a></h3>
<div class="section" id="linux-compilers-from-the-system">
<h4>Linux compilers from the system<a class="headerlink" href="#linux-compilers-from-the-system" title="Permalink to this headline">¶</a></h4>
<p>Installing scikit-learn from source without using conda requires you to have
installed the scikit-learn Python development headers and a working C/C++
compiler with OpenMP support (typically the GCC toolchain).</p>
<p>Install build dependencies for Debian-based operating systems, e.g.
Ubuntu:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">sudo apt-get install build-essential python3-dev python3-pip</span>
</pre></div></div><p>then proceed as usual:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">pip3 install cython</span>
<span class="prompt1">pip3 install --verbose --editable .</span>
</pre></div></div><p>Cython and the pre-compiled wheels for the runtime dependencies (numpy, scipy
and joblib) should automatically be installed in
<code class="docutils literal notranslate"><span class="pre">$HOME/.local/lib/pythonX.Y/site-packages</span></code>. Alternatively you can run the
above commands from a <a class="reference external" href="https://docs.python.org/3/tutorial/venv.html">virtualenv</a> or a <a class="reference external" href="https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html">conda environment</a> to get full
isolation from the Python packages installed via the system packager. When
using an isolated environment, <code class="docutils literal notranslate"><span class="pre">pip3</span></code> should be replaced by <code class="docutils literal notranslate"><span class="pre">pip</span></code> in the
above commands.</p>
<p>When precompiled wheels of the runtime dependencies are not avalaible for your
architecture (e.g. ARM), you can install the system versions:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">sudo apt-get install cython3 python3-numpy python3-scipy</span>
</pre></div></div><p>On Red Hat and clones (e.g. CentOS), install the dependencies using:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">sudo yum -y install gcc gcc-c++ python3-devel numpy scipy</span>
</pre></div></div></div>
<div class="section" id="linux-compilers-from-conda-forge">
<h4>Linux compilers from conda-forge<a class="headerlink" href="#linux-compilers-from-conda-forge" title="Permalink to this headline">¶</a></h4>
<p>Alternatively, install a recent version of the GNU C Compiler toolchain (GCC)
in the user folder using conda:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">conda create -n sklearn-dev -c conda-forge python numpy scipy cython <span class="se">\</span>
joblib threadpoolctl pytest compilers</span>
<span class="prompt1">conda activate sklearn-dev</span>
<span class="prompt1">pip install --verbose --no-build-isolation --editable .</span>
</pre></div></div></div>
</div>
<div class="section" id="freebsd">
<span id="compiler-freebsd"></span><h3>FreeBSD<a class="headerlink" href="#freebsd" title="Permalink to this headline">¶</a></h3>
<p>The clang compiler included in FreeBSD 12.0 and 11.2 base systems does not
include OpenMP support. You need to install the <code class="docutils literal notranslate"><span class="pre">openmp</span></code> library from packages
(or ports):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">sudo pkg install openmp</span>
</pre></div></div><p>This will install header files in <code class="docutils literal notranslate"><span class="pre">/usr/local/include</span></code> and libs in
<code class="docutils literal notranslate"><span class="pre">/usr/local/lib</span></code>. Since these directories are not searched by default, you
can set the environment variables to these locations:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1"><span class="nb">export</span> <span class="nv">CFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$CFLAGS</span><span class="s2"> -I/usr/local/include"</span></span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">CXXFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$CXXFLAGS</span><span class="s2"> -I/usr/local/include"</span></span>
<span class="prompt1"><span class="nb">export</span> <span class="nv">LDFLAGS</span><span class="o">=</span><span class="s2">"</span><span class="nv">$LDFLAGS</span><span class="s2"> -Wl,-rpath,/usr/local/lib -L/usr/local/lib -lomp"</span></span>
</pre></div></div><p>Finally, build the package using the standard command:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">pip install --verbose --no-build-isolation --editable .</span>
</pre></div></div><p>For the upcoming FreeBSD 12.1 and 11.3 versions, OpenMP will be included in
the base system and these steps will not be necessary.</p>
</div>
</div>
<div class="section" id="alternative-compilers">
<h2>Alternative compilers<a class="headerlink" href="#alternative-compilers" title="Permalink to this headline">¶</a></h2>
<p>The command:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">pip install --verbose --editable .</span>
</pre></div></div><p>will build scikit-learn using your default C/C++ compiler. If you want to build
scikit-learn with another compiler handled by <code class="docutils literal notranslate"><span class="pre">distutils</span></code> or by
<code class="docutils literal notranslate"><span class="pre">numpy.distutils</span></code>, use the following command:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">python setup.py build_ext --compiler<span class="o">=</span><compiler> -i build_clib --compiler<span class="o">=</span><compiler></span>
</pre></div></div><p>To see the list of available compilers run:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">python setup.py build_ext --help-compiler</span>
</pre></div></div><p>If your compiler is not listed here, you can specify it via the <code class="docutils literal notranslate"><span class="pre">CC</span></code> and
<code class="docutils literal notranslate"><span class="pre">LDSHARED</span></code> environment variables (does not work on windows):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1"><span class="nv">CC</span><span class="o">=</span><compiler> <span class="nv">LDSHARED</span><span class="o">=</span><span class="s2">"<compiler> -shared"</span> python setup.py build_ext -i</span>
</pre></div></div><div class="section" id="building-with-intel-c-compiler-icc-using-oneapi-on-linux">
<h3>Building with Intel C Compiler (ICC) using oneAPI on Linux<a class="headerlink" href="#building-with-intel-c-compiler-icc-using-oneapi-on-linux" title="Permalink to this headline">¶</a></h3>
<p>Intel provides access to all of its oneAPI toolkits and packages through a
public APT repository. First you need to get and install the public key of this
repository:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">wget https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB</span>
<span class="prompt1">sudo apt-key add GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB</span>
<span class="prompt1">rm GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB</span>
</pre></div></div><p>Then, add the oneAPI repository to your APT repositories:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">sudo add-apt-repository <span class="s2">"deb https://apt.repos.intel.com/oneapi all main"</span></span>
<span class="prompt1">sudo apt-get update</span>
</pre></div></div><p>Install ICC, packaged under the name
<code class="docutils literal notranslate"><span class="pre">intel-oneapi-compiler-dpcpp-cpp-and-cpp-classic</span></code>:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">sudo apt-get install intel-oneapi-compiler-dpcpp-cpp-and-cpp-classic</span>
</pre></div></div><p>Before using ICC, you need to set up environment variables:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1"><span class="nb">source</span> /opt/intel/oneapi/setvars.sh</span>
</pre></div></div><p>Finally, you can build scikit-learn. For example on Linux x86_64:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span class="prompt1">python setup.py build_ext --compiler<span class="o">=</span>intelem -i build_clib --compiler<span class="o">=</span>intelem</span>
</pre></div></div></div>
</div>
<div class="section" id="parallel-builds">
<h2>Parallel builds<a class="headerlink" href="#parallel-builds" title="Permalink to this headline">¶</a></h2>
<p>It is possible to build scikit-learn compiled extensions in parallel by setting
and environment variable as follows before calling the <code class="docutils literal notranslate"><span class="pre">pip</span> <span class="pre">install</span></code> or
<code class="docutils literal notranslate"><span class="pre">python</span> <span class="pre">setup.py</span> <span class="pre">build_ext</span></code> commands:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">export</span> <span class="n">SKLEARN_BUILD_PARALLEL</span><span class="o">=</span><span class="mi">3</span>
<span class="n">pip</span> <span class="n">install</span> <span class="o">--</span><span class="n">verbose</span> <span class="o">--</span><span class="n">no</span><span class="o">-</span><span class="n">build</span><span class="o">-</span><span class="n">isolation</span> <span class="o">--</span><span class="n">editable</span> <span class="o">.</span>
</pre></div>
</div>
<p>On a machine with 2 CPU cores, it can be beneficial to use a parallelism level
of 3 to overlap IO bound tasks (reading and writing files on disk) with CPU
bound tasks (actually compiling).</p>
</div>
</div>
</div>
<div class="container">
<footer class="sk-content-footer">
© 2007 - 2020, scikit-learn developers (BSD License).
<a href="../_sources/developers/advanced_installation.rst.txt" rel="nofollow">Show this page source</a>
</footer>
</div>
</div>
</div>
<script src="../_static/js/vendor/bootstrap.min.js"></script>
<script>
window.ga=window.ga||function(){(ga.q=ga.q||[]).push(arguments)};ga.l=+new Date;
ga('create', 'UA-22606712-2', 'auto');
ga('set', 'anonymizeIp', true);
ga('send', 'pageview');
</script>
<script async src='https://www.google-analytics.com/analytics.js'></script>
<script>
$(document).ready(function() {
/* Add a [>>>] button on the top-right corner of code samples to hide
* the >>> and ... prompts and the output and thus make the code
* copyable. */
var div = $('.highlight-python .highlight,' +
'.highlight-python3 .highlight,' +
'.highlight-pycon .highlight,' +
'.highlight-default .highlight')
var pre = div.find('pre');
// get the styles from the current theme
pre.parent().parent().css('position', 'relative');
var hide_text = 'Hide prompts and outputs';
var show_text = 'Show prompts and outputs';
// create and add the button to all the code blocks that contain >>>
div.each(function(index) {
var jthis = $(this);
if (jthis.find('.gp').length > 0) {
var button = $('<span class="copybutton">>>></span>');
button.attr('title', hide_text);
button.data('hidden', 'false');
jthis.prepend(button);
}
// tracebacks (.gt) contain bare text elements that need to be
// wrapped in a span to work with .nextUntil() (see later)
jthis.find('pre:has(.gt)').contents().filter(function() {
return ((this.nodeType == 3) && (this.data.trim().length > 0));
}).wrap('<span>');
});
// define the behavior of the button when it's clicked
$('.copybutton').click(function(e){
e.preventDefault();
var button = $(this);
if (button.data('hidden') === 'false') {
// hide the code output
button.parent().find('.go, .gp, .gt').hide();
button.next('pre').find('.gt').nextUntil('.gp, .go').css('visibility', 'hidden');
button.css('text-decoration', 'line-through');
button.attr('title', show_text);
button.data('hidden', 'true');
} else {
// show the code output
button.parent().find('.go, .gp, .gt').show();
button.next('pre').find('.gt').nextUntil('.gp, .go').css('visibility', 'visible');
button.css('text-decoration', 'none');
button.attr('title', hide_text);
button.data('hidden', 'false');
}
});
/*** Add permalink buttons next to glossary terms ***/
$('dl.glossary > dt[id]').append(function() {
return ('<a class="headerlink" href="#' +
this.getAttribute('id') +
'" title="Permalink to this term">¶</a>');
});
/*** Hide navbar when scrolling down ***/
// Returns true when headerlink target matches hash in url
(function() {
hashTargetOnTop = function() {
var hash = window.location.hash;
if ( hash.length < 2 ) { return false; }
var target = document.getElementById( hash.slice(1) );
if ( target === null ) { return false; }
var top = target.getBoundingClientRect().top;
return (top < 2) && (top > -2);
};
// Hide navbar on load if hash target is on top
var navBar = document.getElementById("navbar");
var navBarToggler = document.getElementById("sk-navbar-toggler");
var navBarHeightHidden = "-" + navBar.getBoundingClientRect().height + "px";
var $window = $(window);
hideNavBar = function() {
navBar.style.top = navBarHeightHidden;
};
showNavBar = function() {
navBar.style.top = "0";
}
if (hashTargetOnTop()) {
hideNavBar()
}
var prevScrollpos = window.pageYOffset;
hideOnScroll = function(lastScrollTop) {
if (($window.width() < 768) && (navBarToggler.getAttribute("aria-expanded") === 'true')) {
return;
}
if (lastScrollTop > 2 && (prevScrollpos <= lastScrollTop) || hashTargetOnTop()){
hideNavBar()
} else {
showNavBar()
}
prevScrollpos = lastScrollTop;
};
/*** high performance scroll event listener***/
var raf = window.requestAnimationFrame ||
window.webkitRequestAnimationFrame ||
window.mozRequestAnimationFrame ||
window.msRequestAnimationFrame ||
window.oRequestAnimationFrame;
var lastScrollTop = $window.scrollTop();
if (raf) {
loop();
}
function loop() {
var scrollTop = $window.scrollTop();
if (lastScrollTop === scrollTop) {
raf(loop);
return;
} else {
lastScrollTop = scrollTop;
hideOnScroll(lastScrollTop);
raf(loop);
}
}
})();
});
</script>
<script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-chtml.js"></script>
<script src="https://scikit-learn.org/versionwarning.js"></script>
</body>
</html>