-
Notifications
You must be signed in to change notification settings - Fork 3
/
vowpalwabbit.sklearn.html
2217 lines (2175 loc) · 133 KB
/
vowpalwabbit.sklearn.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 PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="X-UA-Compatible" content="IE=Edge" />
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>vowpalwabbit.sklearn — VowpalWabbit 8.10.2 documentation</title>
<link rel="stylesheet" href="_static/classic.css" type="text/css" />
<link rel="stylesheet" href="_static/pygments.css" type="text/css" />
<script type="text/javascript" id="documentation_options" data-url_root="./" src="_static/documentation_options.js"></script>
<script type="text/javascript" src="_static/jquery.js"></script>
<script type="text/javascript" src="_static/underscore.js"></script>
<script type="text/javascript" src="_static/doctools.js"></script>
<script type="text/javascript" src="_static/language_data.js"></script>
<link rel="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
</head><body>
<div class="related" role="navigation" aria-label="related navigation">
<h3>Navigation</h3>
<ul>
<li class="right" style="margin-right: 10px">
<a href="genindex.html" title="General Index"
accesskey="I">index</a></li>
<li class="right" >
<a href="py-modindex.html" title="Python Module Index"
>modules</a> |</li>
<li class="nav-item nav-item-0"><a href="index.html">VowpalWabbit 8.10.2 documentation</a> »</li>
</ul>
</div>
<div class="document">
<div class="documentwrapper">
<div class="bodywrapper">
<div class="body" role="main">
<div class="section" id="module-vowpalwabbit.sklearn_vw">
<span id="vowpalwabbit-sklearn"></span><h1>vowpalwabbit.sklearn<a class="headerlink" href="#module-vowpalwabbit.sklearn_vw" title="Permalink to this headline">¶</a></h1>
<p>Utilities to support integration of Vowpal Wabbit and scikit-learn</p>
<dl class="class">
<dt id="vowpalwabbit.sklearn_vw.LinearClassifierMixin">
<em class="property">class </em><code class="descclassname">vowpalwabbit.sklearn_vw.</code><code class="descname">LinearClassifierMixin</code><a class="headerlink" href="#vowpalwabbit.sklearn_vw.LinearClassifierMixin" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">sklearn.linear_model.logistic.LogisticRegression</span></code></p>
<p class="rubric">Methods</p>
<table border="1" class="longtable docutils">
<colgroup>
<col width="10%" />
<col width="90%" />
</colgroup>
<tbody valign="top">
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">decision_function</span></code>(self, X)</td>
<td>Predict confidence scores for samples.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">densify</span></code>(self)</td>
<td>Convert coefficient matrix to dense array format.</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">fit</span></code>(self, X, y[, sample_weight])</td>
<td>Fit the model according to the given training data.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_params</span></code>(self[, deep])</td>
<td>Get parameters for this estimator.</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict</span></code>(self, X)</td>
<td>Predict class labels for samples in X.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict_log_proba</span></code>(self, X)</td>
<td>Log of probability estimates.</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict_proba</span></code>(self, X)</td>
<td>Probability estimates.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">score</span></code>(self, X, y[, sample_weight])</td>
<td>Returns the mean accuracy on the given test data and labels.</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">set_params</span></code>(self, \*\*params)</td>
<td>Set the parameters of this estimator.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">sparsify</span></code>(self)</td>
<td>Convert coefficient matrix to sparse format.</td>
</tr>
</tbody>
</table>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.LinearClassifierMixin.__init__">
<code class="descname">__init__</code><span class="sig-paren">(</span><em>self</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.LinearClassifierMixin.__init__" title="Permalink to this definition">¶</a></dt>
<dd><p>x.__init__(…) initializes x; see help(type(x)) for signature</p>
</dd></dl>
</dd></dl>
<dl class="class">
<dt id="vowpalwabbit.sklearn_vw.VW">
<em class="property">class </em><code class="descclassname">vowpalwabbit.sklearn_vw.</code><code class="descname">VW</code><span class="sig-paren">(</span><em>convert_to_vw=True</em>, <em>convert_labels=True</em>, <em>ring_size=None</em>, <em>strict_parse=None</em>, <em>learning_rate=None</em>, <em>l=None</em>, <em>power_t=None</em>, <em>decay_learning_rate=None</em>, <em>initial_t=None</em>, <em>feature_mask=None</em>, <em>initial_regressor=None</em>, <em>i=None</em>, <em>initial_weight=None</em>, <em>random_weights=None</em>, <em>normal_weights=None</em>, <em>truncated_normal_weights=None</em>, <em>sparse_weights=None</em>, <em>input_feature_regularizer=None</em>, <em>quiet=True</em>, <em>random_seed=None</em>, <em>hash=None</em>, <em>hash_seed=None</em>, <em>ignore=None</em>, <em>ignore_linear=None</em>, <em>keep=None</em>, <em>redefine=None</em>, <em>bit_precision=None</em>, <em>b=None</em>, <em>noconstant=None</em>, <em>constant=None</em>, <em>C=None</em>, <em>ngram=None</em>, <em>skips=None</em>, <em>feature_limit=None</em>, <em>affix=None</em>, <em>spelling=None</em>, <em>dictionary=None</em>, <em>dictionary_path=None</em>, <em>interactions=None</em>, <em>permutations=None</em>, <em>leave_duplicate_interactions=None</em>, <em>quadratic=None</em>, <em>q=None</em>, <em>cubic=None</em>, <em>testonly=None</em>, <em>t=None</em>, <em>holdout_off=None</em>, <em>holdout_period=None</em>, <em>holdout_after=None</em>, <em>early_terminate=None</em>, <em>passes=1</em>, <em>initial_pass_length=None</em>, <em>examples=None</em>, <em>min_prediction=None</em>, <em>max_prediction=None</em>, <em>sort_features=None</em>, <em>loss_function=None</em>, <em>quantile_tau=None</em>, <em>l1=None</em>, <em>l2=None</em>, <em>no_bias_regularization=None</em>, <em>named_labels=None</em>, <em>final_regressor=None</em>, <em>f=None</em>, <em>readable_model=None</em>, <em>invert_hash=None</em>, <em>save_resume=None</em>, <em>preserve_performance_counters=None</em>, <em>output_feature_regularizer_binary=None</em>, <em>output_feature_regularizer_text=None</em>, <em>oaa=None</em>, <em>ect=None</em>, <em>csoaa=None</em>, <em>wap=None</em>, <em>probabilities=None</em>, <em>nn=None</em>, <em>inpass=None</em>, <em>multitask=None</em>, <em>dropout=None</em>, <em>meanfield=None</em>, <em>conjugate_gradient=None</em>, <em>bfgs=None</em>, <em>hessian_on=None</em>, <em>mem=None</em>, <em>termination=None</em>, <em>lda=None</em>, <em>lda_alpha=None</em>, <em>lda_rho=None</em>, <em>lda_D=None</em>, <em>lda_epsilon=None</em>, <em>minibatch=None</em>, <em>svrg=None</em>, <em>stage_size=None</em>, <em>ftrl=None</em>, <em>coin=None</em>, <em>pistol=None</em>, <em>ftrl_alpha=None</em>, <em>ftrl_beta=None</em>, <em>ksvm=None</em>, <em>kernel=None</em>, <em>bandwidth=None</em>, <em>degree=None</em>, <em>sgd=None</em>, <em>adaptive=None</em>, <em>invariant=None</em>, <em>normalized=None</em>, <em>link=None</em>, <em>stage_poly=None</em>, <em>sched_exponent=None</em>, <em>batch_sz=None</em>, <em>batch_sz_no_doubling=None</em>, <em>lrq=None</em>, <em>lrqdropout=None</em>, <em>lrqfa=None</em>, <em>data=None</em>, <em>d=None</em>, <em>cache=None</em>, <em>c=None</em>, <em>cache_file=None</em>, <em>json=None</em>, <em>kill_cache=None</em>, <em>k=None</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">sklearn.base.BaseEstimator</span></code></p>
<p>Vowpal Wabbit Scikit-learn Base Estimator wrapper</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Attributes:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>convert_to_vw</strong> <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">flag to convert X input to vw format</p>
</dd>
<dt><strong>convert_labels</strong> <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Convert labels of the form [0,1] to [-1,1]</p>
</dd>
<dt><strong>vw_</strong> <span class="classifier-delimiter">:</span> <span class="classifier">pyvw.vw</span></dt>
<dd><p class="first last">vw instance</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<p class="rubric">Methods</p>
<table border="1" class="longtable docutils">
<colgroup>
<col width="10%" />
<col width="90%" />
</colgroup>
<tbody valign="top">
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.fit" title="vowpalwabbit.sklearn_vw.VW.fit"><code class="xref py py-obj docutils literal notranslate"><span class="pre">fit</span></code></a>(self[, X, y, sample_weight])</td>
<td>Fit the model according to the given training data</td>
</tr>
<tr class="row-even"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.get_coefs" title="vowpalwabbit.sklearn_vw.VW.get_coefs"><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_coefs</span></code></a>(self)</td>
<td>Returns coefficient weights as ordered sparse matrix</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.get_intercept" title="vowpalwabbit.sklearn_vw.VW.get_intercept"><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_intercept</span></code></a>(self)</td>
<td>Returns intercept weight for model</td>
</tr>
<tr class="row-even"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.get_params" title="vowpalwabbit.sklearn_vw.VW.get_params"><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_params</span></code></a>(self[, deep])</td>
<td>This returns the full set of vw and estimator parameters currently in use</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.get_vw" title="vowpalwabbit.sklearn_vw.VW.get_vw"><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_vw</span></code></a>(self)</td>
<td>Get the vw instance</td>
</tr>
<tr class="row-even"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.load" title="vowpalwabbit.sklearn_vw.VW.load"><code class="xref py py-obj docutils literal notranslate"><span class="pre">load</span></code></a>(self, filename)</td>
<td>Load model from file</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.predict" title="vowpalwabbit.sklearn_vw.VW.predict"><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict</span></code></a>(self, X)</td>
<td>Predict with Vowpal Wabbit model</td>
</tr>
<tr class="row-even"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.save" title="vowpalwabbit.sklearn_vw.VW.save"><code class="xref py py-obj docutils literal notranslate"><span class="pre">save</span></code></a>(self, filename)</td>
<td>Save model to file</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.set_coefs" title="vowpalwabbit.sklearn_vw.VW.set_coefs"><code class="xref py py-obj docutils literal notranslate"><span class="pre">set_coefs</span></code></a>(self, coefs)</td>
<td>Sets coefficients weights from ordered sparse matrix</td>
</tr>
<tr class="row-even"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW.set_params" title="vowpalwabbit.sklearn_vw.VW.set_params"><code class="xref py py-obj docutils literal notranslate"><span class="pre">set_params</span></code></a>(self, \*\*kwargs)</td>
<td>This destroys and recreates the Vowpal Wabbit model with updated parameters any parameters not provided will remain as they are currently</td>
</tr>
</tbody>
</table>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.__init__">
<code class="descname">__init__</code><span class="sig-paren">(</span><em>self</em>, <em>convert_to_vw=True</em>, <em>convert_labels=True</em>, <em>ring_size=None</em>, <em>strict_parse=None</em>, <em>learning_rate=None</em>, <em>l=None</em>, <em>power_t=None</em>, <em>decay_learning_rate=None</em>, <em>initial_t=None</em>, <em>feature_mask=None</em>, <em>initial_regressor=None</em>, <em>i=None</em>, <em>initial_weight=None</em>, <em>random_weights=None</em>, <em>normal_weights=None</em>, <em>truncated_normal_weights=None</em>, <em>sparse_weights=None</em>, <em>input_feature_regularizer=None</em>, <em>quiet=True</em>, <em>random_seed=None</em>, <em>hash=None</em>, <em>hash_seed=None</em>, <em>ignore=None</em>, <em>ignore_linear=None</em>, <em>keep=None</em>, <em>redefine=None</em>, <em>bit_precision=None</em>, <em>b=None</em>, <em>noconstant=None</em>, <em>constant=None</em>, <em>C=None</em>, <em>ngram=None</em>, <em>skips=None</em>, <em>feature_limit=None</em>, <em>affix=None</em>, <em>spelling=None</em>, <em>dictionary=None</em>, <em>dictionary_path=None</em>, <em>interactions=None</em>, <em>permutations=None</em>, <em>leave_duplicate_interactions=None</em>, <em>quadratic=None</em>, <em>q=None</em>, <em>cubic=None</em>, <em>testonly=None</em>, <em>t=None</em>, <em>holdout_off=None</em>, <em>holdout_period=None</em>, <em>holdout_after=None</em>, <em>early_terminate=None</em>, <em>passes=1</em>, <em>initial_pass_length=None</em>, <em>examples=None</em>, <em>min_prediction=None</em>, <em>max_prediction=None</em>, <em>sort_features=None</em>, <em>loss_function=None</em>, <em>quantile_tau=None</em>, <em>l1=None</em>, <em>l2=None</em>, <em>no_bias_regularization=None</em>, <em>named_labels=None</em>, <em>final_regressor=None</em>, <em>f=None</em>, <em>readable_model=None</em>, <em>invert_hash=None</em>, <em>save_resume=None</em>, <em>preserve_performance_counters=None</em>, <em>output_feature_regularizer_binary=None</em>, <em>output_feature_regularizer_text=None</em>, <em>oaa=None</em>, <em>ect=None</em>, <em>csoaa=None</em>, <em>wap=None</em>, <em>probabilities=None</em>, <em>nn=None</em>, <em>inpass=None</em>, <em>multitask=None</em>, <em>dropout=None</em>, <em>meanfield=None</em>, <em>conjugate_gradient=None</em>, <em>bfgs=None</em>, <em>hessian_on=None</em>, <em>mem=None</em>, <em>termination=None</em>, <em>lda=None</em>, <em>lda_alpha=None</em>, <em>lda_rho=None</em>, <em>lda_D=None</em>, <em>lda_epsilon=None</em>, <em>minibatch=None</em>, <em>svrg=None</em>, <em>stage_size=None</em>, <em>ftrl=None</em>, <em>coin=None</em>, <em>pistol=None</em>, <em>ftrl_alpha=None</em>, <em>ftrl_beta=None</em>, <em>ksvm=None</em>, <em>kernel=None</em>, <em>bandwidth=None</em>, <em>degree=None</em>, <em>sgd=None</em>, <em>adaptive=None</em>, <em>invariant=None</em>, <em>normalized=None</em>, <em>link=None</em>, <em>stage_poly=None</em>, <em>sched_exponent=None</em>, <em>batch_sz=None</em>, <em>batch_sz_no_doubling=None</em>, <em>lrq=None</em>, <em>lrqdropout=None</em>, <em>lrqfa=None</em>, <em>data=None</em>, <em>d=None</em>, <em>cache=None</em>, <em>c=None</em>, <em>cache_file=None</em>, <em>json=None</em>, <em>kill_cache=None</em>, <em>k=None</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.__init__" title="Permalink to this definition">¶</a></dt>
<dd><p>VW model constructor, exposing all supported parameters to keep sklearn happy</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><dl class="first docutils">
<dt><strong>Estimator options</strong></dt>
<dd><dl class="first docutils">
<dt>convert_to_vw <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">flag to convert X input to vw format</p>
</dd>
<dt>convert_labels <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Convert labels of the form [0,1] to [-1,1]</p>
</dd>
</dl>
<p>VW options</p>
<dl class="docutils">
<dt>ring_size <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">size of example ring</p>
</dd>
<dt>strict_parse <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">throw on malformed examples</p>
</dd>
</dl>
<p>Update options</p>
<dl class="docutils">
<dt>learning_rate,l <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set learning rate</p>
</dd>
<dt>power_t <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">t power value</p>
</dd>
<dt>decay_learning_rate <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set Decay factor for learning_rate between passes</p>
</dd>
<dt>initial_t <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">initial t value</p>
</dd>
<dt>feature_mask <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Use existing regressor to determine which parameters may be updated.
If no initial_regressor given, also used for initial weights.</p>
</dd>
</dl>
<p>Weight options</p>
<dl class="docutils">
<dt>initial_regressor,i <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Initial regressor(s)</p>
</dd>
<dt>initial_weight <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set all weights to an initial value of arg.</p>
</dd>
<dt>random_weights <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">make initial weights random</p>
</dd>
<dt>normal_weights <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">make initial weights normal</p>
</dd>
<dt>truncated_normal_weights <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">make initial weights truncated normal</p>
</dd>
<dt>sparse_weights <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Use a sparse datastructure for weights</p>
</dd>
<dt>input_feature_regularizer <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Per feature regularization input file</p>
</dd>
</dl>
<p>Diagnostic options</p>
<dl class="docutils">
<dt>quiet <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Don’t output disgnostics and progress updates</p>
</dd>
</dl>
<p>Randomization options</p>
<dl class="docutils">
<dt>random_seed <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">seed random number generator</p>
</dd>
</dl>
<p>Feature options</p>
<dl class="docutils">
<dt>hash <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">how to hash the features. Available options: strings, all</p>
</dd>
<dt>hash_seed <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">seed for hash function</p>
</dd>
<dt>ignore <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">ignore namespaces beginning with character <arg></p>
</dd>
<dt>ignore_linear <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">ignore namespaces beginning with character <arg> for linear terms only</p>
</dd>
<dt>keep <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">keep namespaces beginning with character <arg></p>
</dd>
<dt>redefine <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Redefine namespaces beginning with characters of string S as namespace N. <arg> shall be in
form ‘N:=S’ where := is operator. Empty N or S are treated as default namespace.
Use ‘:’ as a wildcard in S.</p>
</dd>
<dt>bit_precision,b <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">number of bits in the feature table</p>
</dd>
<dt>noconstant <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Don’t add a constant feature</p>
</dd>
<dt>constant,C <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set initial value of constant</p>
</dd>
<dt>ngram <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Generate N grams. To generate N grams for a single namespace ‘foo’, arg should be fN.</p>
</dd>
<dt>skips <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Generate skips in N grams. This in conjunction with the ngram tag can be used to generate
generalized n-skip-k-gram. To generate n-skips for a single namespace ‘foo’, arg should be fN.</p>
</dd>
<dt>feature_limit <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">limit to N features. To apply to a single namespace ‘foo’, arg should be fN</p>
</dd>
<dt>affix <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">generate prefixes/suffixes of features; argument ‘+2a,-3b,+1’ means generate 2-char prefixes for
namespace a, 3-char suffixes for b and 1 char prefixes for default namespace</p>
</dd>
<dt>spelling <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">compute spelling features for a give namespace (use ‘_’ for default namespace)</p>
</dd>
<dt>dictionary <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">read a dictionary for additional features (arg either ‘x:file’ or just ‘file’)</p>
</dd>
<dt>dictionary_path <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">look in this directory for dictionaries; defaults to current directory or env{PATH}</p>
</dd>
<dt>interactions <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Create feature interactions of any level between namespaces.</p>
</dd>
<dt>permutations <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Use permutations instead of combinations for feature interactions of same namespace.</p>
</dd>
<dt>leave_duplicate_interactions <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Don’t remove interactions with duplicate combinations of namespaces. For
ex. this is a duplicate: ‘-q ab -q ba’ and a lot more in ‘-q ::’.</p>
</dd>
<dt>quadratic,q <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Create and use quadratic features, q:: corresponds to a wildcard for all printable characters</p>
</dd>
<dt>cubic <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Create and use cubic features</p>
</dd>
</dl>
<p>Example options</p>
<dl class="docutils">
<dt>testonly,t <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Ignore label information and just test</p>
</dd>
<dt>holdout_off <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">no holdout data in multiple passes</p>
</dd>
<dt>holdout_period <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">holdout period for test only</p>
</dd>
<dt>holdout_after <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">holdout after n training examples</p>
</dd>
<dt>early_terminate <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">Specify the number of passes tolerated when holdout loss doesn’t
decrease before early termination</p>
</dd>
<dt>passes <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">Number of Training Passes</p>
</dd>
<dt>initial_pass_length <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">initial number of examples per pass</p>
</dd>
<dt>examples <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">number of examples to parse</p>
</dd>
<dt>min_prediction <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Smallest prediction to output</p>
</dd>
<dt>max_prediction <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Largest prediction to output</p>
</dd>
<dt>sort_features <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">turn this on to disregard order in which features have been defined. This will lead to
smaller cache sizes</p>
</dd>
<dt>loss_function <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">default_value(“squared”), “Specify the loss function to be used, uses squared by default.
Currently available ones are squared, classic, hinge, logistic and quantile.</p>
</dd>
<dt>quantile_tau <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Parameter tau associated with Quantile loss. Defaults to 0.5</p>
</dd>
<dt>l1 <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">l_1 lambda (L1 regularization)</p>
</dd>
<dt>l2 <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">l_2 lambda (L2 regularization)</p>
</dd>
<dt>no_bias_regularization <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">no bias in regularization</p>
</dd>
<dt>named_labels <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">use names for labels (multiclass, etc.) rather than integers, argument specified all
possible labels, comma-sep, eg “–named_labels Noun,Verb,Adj,Punc”</p>
</dd>
</dl>
<p>Output model</p>
<dl class="last docutils">
<dt>final_regressor,f <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Final regressor</p>
</dd>
<dt>readable_model <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Output human-readable final regressor with numeric features</p>
</dd>
<dt>invert_hash <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Output human-readable final regressor with feature names. Computationally expensive.</p>
</dd>
<dt>save_resume <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">save extra state so learning can be resumed later with new data</p>
</dd>
<dt>preserve_performance_counters <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">reset performance counters when warmstarting</p>
</dd>
<dt>output_feature_regularizer_binary <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Per feature regularization output file</p>
</dd>
<dt>output_feature_regularizer_text <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Per feature regularization output file, in text</p>
</dd>
</dl>
</dd>
<dt><strong>Multiclass options</strong></dt>
<dd><dl class="first docutils">
<dt>oaa <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">Use one-against-all multiclass learning with labels</p>
</dd>
<dt>oaa_subsample <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">subsample this number of negative examples when learning</p>
</dd>
<dt>ect <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">Use error correcting tournament multiclass learning</p>
</dd>
<dt>csoaa <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">Use cost sensitive one-against-all multiclass learning</p>
</dd>
<dt>wap <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">Use weighted all pairs multiclass learning</p>
</dd>
<dt>probabilities <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">predict probabilities of all classes</p>
</dd>
</dl>
<p>Neural Network options</p>
<dl class="docutils">
<dt>nn <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">Use a sigmoidal feed-forward neural network with N hidden units</p>
</dd>
<dt>inpass <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Train or test sigmoidal feed-forward network with input pass-through</p>
</dd>
<dt>multitask <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Share hidden layer across all reduced tasks</p>
</dd>
<dt>dropout <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Train or test sigmoidal feed-forward network using dropout</p>
</dd>
<dt>meanfield <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Train or test sigmoidal feed-forward network using mean field</p>
</dd>
</dl>
<p>LBFGS and Conjugate Gradient options</p>
<dl class="docutils">
<dt>conjugate_gradient <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use conjugate gradient based optimization</p>
</dd>
<dt>bgfs <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use bfgs updates</p>
</dd>
<dt>hessian_on <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use second derivative in line search</p>
</dd>
<dt>mem <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">memory in bfgs</p>
</dd>
<dt>termination <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">termination threshold</p>
</dd>
</dl>
<p>Latent Dirichlet Allocation options</p>
<dl class="docutils">
<dt>lda <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">Run lda with <int> topics</p>
</dd>
<dt>lda_alpha <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Prior on sparsity of per-document topic weights</p>
</dd>
<dt>lda_rho <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Prior on sparsity of topic distributions</p>
</dd>
<dt>lda_D <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">Number of documents</p>
</dd>
<dt>lda_epsilon <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Loop convergence threshold</p>
</dd>
<dt>minibatch <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">Minibatch size for LDA</p>
</dd>
</dl>
<p>Stochastic Variance Reduced Gradient options</p>
<dl class="docutils">
<dt>svrg <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Streaming Stochastic Variance Reduced Gradient</p>
</dd>
<dt>stage_size <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">Number of passes per SVRG stage</p>
</dd>
</dl>
<p>Follow the Regularized Leader options</p>
<dl class="docutils">
<dt>ftrl <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Run Follow the Proximal Regularized Leader</p>
</dd>
<dt>coin <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Coin betting optimizer</p>
</dd>
<dt>pistol <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">PiSTOL: Parameter free STOchastic Learning</p>
</dd>
<dt>ftrl_alpha <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Alpha parameter for FTRL optimization</p>
</dd>
<dt>ftrl_beta <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Beta parameters for FTRL optimization</p>
</dd>
</dl>
<p>Kernel SVM options</p>
<dl class="docutils">
<dt>ksvm <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">kernel svm</p>
</dd>
<dt>kernel <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">type of kernel (rbf or linear (default))</p>
</dd>
<dt>bandwidth <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">bandwidth of rbf kernel</p>
</dd>
<dt>degree <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">degree of poly kernel</p>
</dd>
</dl>
<p>Gradient Descent options</p>
<dl class="docutils">
<dt>sgd <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use regular stochastic gradient descent update</p>
</dd>
<dt>adaptive <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use adaptive, individual learning rates</p>
</dd>
<dt>adax <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use adaptive learning rates with x^2 instead of g^2x^2</p>
</dd>
<dt>invariant <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use save/importance aware updates</p>
</dd>
<dt>normalized <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use per feature normalized updates</p>
</dd>
</dl>
<p>Scorer options</p>
<dl class="docutils">
<dt>link <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Specify the link function: identity, logistic, glf1 or poisson</p>
</dd>
</dl>
<p>Stagewise polynomial options:</p>
<dl class="docutils">
<dt>stage_poly <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use stagewise polynomial feature learning</p>
</dd>
<dt>sched_exponent <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">exponent controlling quantity of included features</p>
</dd>
<dt>batch_sz <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">multiplier on batch size before including more features</p>
</dd>
<dt>batch_sz_no_doubling <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">batch_sz does not double</p>
</dd>
</dl>
<p>Low Rank Quadratics options:</p>
<dl class="docutils">
<dt>lrq <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use low rank quadratic features</p>
</dd>
<dt>lrqdropout <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use dropout training for low rank quadratic features</p>
</dd>
<dt>lrqfa <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">use low rank quadratic features with field aware weights</p>
</dd>
</dl>
<p>Input options</p>
<dl class="last docutils">
<dt>data,d <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">path to data file for fitting external to sklearn</p>
</dd>
<dt>cache,c <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">use a cache. default is <data>.cache</p>
</dd>
<dt>cache_file <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">path to cache file to use</p>
</dd>
<dt>json <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">enable JSON parsing</p>
</dd>
<dt>kill_cache, k <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">do not reuse existing cache file, create a new one always</p>
</dd>
</dl>
</dd>
</dl>
</td>
</tr>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>self</strong> <span class="classifier-delimiter">:</span> <span class="classifier">BaseEstimator</span></dt>
<dd></dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="attribute">
<dt id="vowpalwabbit.sklearn_vw.VW.convert_labels">
<code class="descname">convert_labels</code><em class="property"> = True</em><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.convert_labels" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="attribute">
<dt id="vowpalwabbit.sklearn_vw.VW.convert_to_vw">
<code class="descname">convert_to_vw</code><em class="property"> = True</em><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.convert_to_vw" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>self</em>, <em>X=None</em>, <em>y=None</em>, <em>sample_weight=None</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>Fit the model according to the given training data</p>
<dl class="docutils">
<dt>TODO: for first pass create and store example objects.</dt>
<dd>for N-1 passes use example objects directly (simulate cache file…but in memory for faster processing)</dd>
</dl>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><dl class="first docutils">
<dt><strong>X</strong> <span class="classifier-delimiter">:</span> <span class="classifier">{array-like, sparse matrix}, shape (n_samples, n_features or 1 if not convert_to_vw) or</span></dt>
<dd><p class="first last">Training vector, where n_samples in the number of samples and
n_features is the number of features.
if not using convert_to_vw, X is expected to be a list of vw formatted feature vector strings with labels</p>
</dd>
<dt><strong>y</strong> <span class="classifier-delimiter">:</span> <span class="classifier">array-like, shape (n_samples,), optional if not convert_to_vw</span></dt>
<dd><p class="first last">Target vector relative to X.</p>
</dd>
<dt><strong>sample_weight</strong> <span class="classifier-delimiter">:</span> <span class="classifier">array-like, shape (n_samples,)</span></dt>
<dd><p class="first last">sample weight vector relative to X.</p>
</dd>
</dl>
</td>
</tr>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>self</strong> <span class="classifier-delimiter">:</span> <span class="classifier">BaseEstimator</span></dt>
<dd><p class="first last">So pipeline can call transform() after fit</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.get_coefs">
<code class="descname">get_coefs</code><span class="sig-paren">(</span><em>self</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.get_coefs" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns coefficient weights as ordered sparse matrix</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Returns:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>sparse matrix</strong> <span class="classifier-delimiter">:</span> <span class="classifier">coefficient weights for model</span></dt>
<dd></dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.get_intercept">
<code class="descname">get_intercept</code><span class="sig-paren">(</span><em>self</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.get_intercept" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns intercept weight for model</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Returns:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>intercept value</strong> <span class="classifier-delimiter">:</span> <span class="classifier">integer, 0 if no constant</span></dt>
<dd></dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.get_params">
<code class="descname">get_params</code><span class="sig-paren">(</span><em>self</em>, <em>deep=True</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.get_params" title="Permalink to this definition">¶</a></dt>
<dd><p>This returns the full set of vw and estimator parameters currently in use</p>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.get_vw">
<code class="descname">get_vw</code><span class="sig-paren">(</span><em>self</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.get_vw" title="Permalink to this definition">¶</a></dt>
<dd><p>Get the vw instance</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Returns:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>vw</strong> <span class="classifier-delimiter">:</span> <span class="classifier">pyvw.vw instance</span></dt>
<dd></dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.load">
<code class="descname">load</code><span class="sig-paren">(</span><em>self</em>, <em>filename</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.load" title="Permalink to this definition">¶</a></dt>
<dd><p>Load model from file</p>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.predict">
<code class="descname">predict</code><span class="sig-paren">(</span><em>self</em>, <em>X</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.predict" title="Permalink to this definition">¶</a></dt>
<dd><p>Predict with Vowpal Wabbit model</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><dl class="first docutils">
<dt><strong>X</strong> <span class="classifier-delimiter">:</span> <span class="classifier">{array-like, sparse matrix}, shape (n_samples, n_features or 1)</span></dt>
<dd><p class="first last">Training vector, where n_samples in the number of samples and
n_features is the number of features.
if not using convert_to_vw, X is expected to be a list of vw formatted feature vector strings with labels</p>
</dd>
</dl>
</td>
</tr>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>y</strong> <span class="classifier-delimiter">:</span> <span class="classifier">array-like, shape (n_samples, 1 or n_classes)</span></dt>
<dd><p class="first last">Output vector relative to X.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.save">
<code class="descname">save</code><span class="sig-paren">(</span><em>self</em>, <em>filename</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.save" title="Permalink to this definition">¶</a></dt>
<dd><p>Save model to file</p>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.set_coefs">
<code class="descname">set_coefs</code><span class="sig-paren">(</span><em>self</em>, <em>coefs</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.set_coefs" title="Permalink to this definition">¶</a></dt>
<dd><p>Sets coefficients weights from ordered sparse matrix</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>coefs</strong> <span class="classifier-delimiter">:</span> <span class="classifier">sparse matrix</span></dt>
<dd><p class="first last">coefficient weights for model</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VW.set_params">
<code class="descname">set_params</code><span class="sig-paren">(</span><em>self</em>, <em>**kwargs</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.set_params" title="Permalink to this definition">¶</a></dt>
<dd><p>This destroys and recreates the Vowpal Wabbit model with updated parameters
any parameters not provided will remain as they are currently</p>
</dd></dl>
<dl class="attribute">
<dt id="vowpalwabbit.sklearn_vw.VW.vw_">
<code class="descname">vw_</code><em class="property"> = None</em><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VW.vw_" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
</dd></dl>
<dl class="class">
<dt id="vowpalwabbit.sklearn_vw.VWClassifier">
<em class="property">class </em><code class="descclassname">vowpalwabbit.sklearn_vw.</code><code class="descname">VWClassifier</code><span class="sig-paren">(</span><em>loss_function='logistic'</em>, <em>**kwargs</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VWClassifier" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <a class="reference internal" href="#vowpalwabbit.sklearn_vw.VW" title="vowpalwabbit.sklearn_vw.VW"><code class="xref py py-class docutils literal notranslate"><span class="pre">vowpalwabbit.sklearn_vw.VW</span></code></a>, <a class="reference internal" href="#vowpalwabbit.sklearn_vw.LinearClassifierMixin" title="vowpalwabbit.sklearn_vw.LinearClassifierMixin"><code class="xref py py-class docutils literal notranslate"><span class="pre">vowpalwabbit.sklearn_vw.LinearClassifierMixin</span></code></a></p>
<p>Vowpal Wabbit Classifier model for binary classification
Use VWMultiClassifier for multiclass classification
Note - We are assuming the VW.predict returns logits, applying link=logistic will break this assumption</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Attributes:</th><td class="field-body"><dl class="first last docutils">
<dt><strong>coef_</strong> <span class="classifier-delimiter">:</span> <span class="classifier">scipy.sparse_matrix</span></dt>
<dd><p class="first last">Empty sparse matrix used the check if model has been fit</p>
</dd>
<dt><strong>classes_</strong> <span class="classifier-delimiter">:</span> <span class="classifier">np.array</span></dt>
<dd><p class="first last">Binary class labels</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<p class="rubric">Methods</p>
<table border="1" class="longtable docutils">
<colgroup>
<col width="10%" />
<col width="90%" />
</colgroup>
<tbody valign="top">
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VWClassifier.decision_function" title="vowpalwabbit.sklearn_vw.VWClassifier.decision_function"><code class="xref py py-obj docutils literal notranslate"><span class="pre">decision_function</span></code></a>(self, X)</td>
<td>Predict confidence scores for samples.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">densify</span></code>(self)</td>
<td>Convert coefficient matrix to dense array format.</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VWClassifier.fit" title="vowpalwabbit.sklearn_vw.VWClassifier.fit"><code class="xref py py-obj docutils literal notranslate"><span class="pre">fit</span></code></a>(self[, X, y, sample_weight])</td>
<td>Fit the model according to the given training data.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_coefs</span></code>(self)</td>
<td>Returns coefficient weights as ordered sparse matrix</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_intercept</span></code>(self)</td>
<td>Returns intercept weight for model</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_params</span></code>(self[, deep])</td>
<td>This returns the full set of vw and estimator parameters currently in use</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_vw</span></code>(self)</td>
<td>Get the vw instance</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">load</span></code>(self, filename)</td>
<td>Load model from file</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VWClassifier.predict" title="vowpalwabbit.sklearn_vw.VWClassifier.predict"><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict</span></code></a>(self, X)</td>
<td>Predict class labels for samples in X.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict_log_proba</span></code>(self, X)</td>
<td>Log of probability estimates.</td>
</tr>
<tr class="row-odd"><td><a class="reference internal" href="#vowpalwabbit.sklearn_vw.VWClassifier.predict_proba" title="vowpalwabbit.sklearn_vw.VWClassifier.predict_proba"><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict_proba</span></code></a>(self, X)</td>
<td>Predict probabilities for samples</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">save</span></code>(self, filename)</td>
<td>Save model to file</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">score</span></code>(self, X, y[, sample_weight])</td>
<td>Returns the mean accuracy on the given test data and labels.</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">set_coefs</span></code>(self, coefs)</td>
<td>Sets coefficients weights from ordered sparse matrix</td>
</tr>
<tr class="row-odd"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">set_params</span></code>(self, \*\*kwargs)</td>
<td>This destroys and recreates the Vowpal Wabbit model with updated parameters any parameters not provided will remain as they are currently</td>
</tr>
<tr class="row-even"><td><code class="xref py py-obj docutils literal notranslate"><span class="pre">sparsify</span></code>(self)</td>
<td>Convert coefficient matrix to sparse format.</td>
</tr>
</tbody>
</table>
<dl class="method">
<dt id="vowpalwabbit.sklearn_vw.VWClassifier.__init__">
<code class="descname">__init__</code><span class="sig-paren">(</span><em>self</em>, <em>loss_function='logistic'</em>, <em>**kwargs</em><span class="sig-paren">)</span><a class="headerlink" href="#vowpalwabbit.sklearn_vw.VWClassifier.__init__" title="Permalink to this definition">¶</a></dt>
<dd><p>VW model constructor, exposing all supported parameters to keep sklearn happy</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><dl class="first docutils">
<dt><strong>Estimator options</strong></dt>
<dd><dl class="first docutils">
<dt>convert_to_vw <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">flag to convert X input to vw format</p>
</dd>
<dt>convert_labels <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Convert labels of the form [0,1] to [-1,1]</p>
</dd>
</dl>
<p>VW options</p>
<dl class="docutils">
<dt>ring_size <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">size of example ring</p>
</dd>
<dt>strict_parse <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">throw on malformed examples</p>
</dd>
</dl>
<p>Update options</p>
<dl class="docutils">
<dt>learning_rate,l <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set learning rate</p>
</dd>
<dt>power_t <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">t power value</p>
</dd>
<dt>decay_learning_rate <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set Decay factor for learning_rate between passes</p>
</dd>
<dt>initial_t <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">initial t value</p>
</dd>
<dt>feature_mask <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Use existing regressor to determine which parameters may be updated.
If no initial_regressor given, also used for initial weights.</p>
</dd>
</dl>
<p>Weight options</p>
<dl class="docutils">
<dt>initial_regressor,i <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Initial regressor(s)</p>
</dd>
<dt>initial_weight <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set all weights to an initial value of arg.</p>
</dd>
<dt>random_weights <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">make initial weights random</p>
</dd>
<dt>normal_weights <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">make initial weights normal</p>
</dd>
<dt>truncated_normal_weights <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">make initial weights truncated normal</p>
</dd>
<dt>sparse_weights <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Use a sparse datastructure for weights</p>
</dd>
<dt>input_feature_regularizer <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Per feature regularization input file</p>
</dd>
</dl>
<p>Diagnostic options</p>
<dl class="docutils">
<dt>quiet <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Don’t output disgnostics and progress updates</p>
</dd>
</dl>
<p>Randomization options</p>
<dl class="docutils">
<dt>random_seed <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">seed random number generator</p>
</dd>
</dl>
<p>Feature options</p>
<dl class="docutils">
<dt>hash <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">how to hash the features. Available options: strings, all</p>
</dd>
<dt>hash_seed <span class="classifier-delimiter">:</span> <span class="classifier">int</span></dt>
<dd><p class="first last">seed for hash function</p>
</dd>
<dt>ignore <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">ignore namespaces beginning with character <arg></p>
</dd>
<dt>ignore_linear <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">ignore namespaces beginning with character <arg> for linear terms only</p>
</dd>
<dt>keep <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">keep namespaces beginning with character <arg></p>
</dd>
<dt>redefine <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Redefine namespaces beginning with characters of string S as namespace N. <arg> shall be in
form ‘N:=S’ where := is operator. Empty N or S are treated as default namespace.
Use ‘:’ as a wildcard in S.</p>
</dd>
<dt>bit_precision,b <span class="classifier-delimiter">:</span> <span class="classifier">integer</span></dt>
<dd><p class="first last">number of bits in the feature table</p>
</dd>
<dt>noconstant <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Don’t add a constant feature</p>
</dd>
<dt>constant,C <span class="classifier-delimiter">:</span> <span class="classifier">float</span></dt>
<dd><p class="first last">Set initial value of constant</p>
</dd>
<dt>ngram <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Generate N grams. To generate N grams for a single namespace ‘foo’, arg should be fN.</p>
</dd>
<dt>skips <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Generate skips in N grams. This in conjunction with the ngram tag can be used to generate
generalized n-skip-k-gram. To generate n-skips for a single namespace ‘foo’, arg should be fN.</p>
</dd>
<dt>feature_limit <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">limit to N features. To apply to a single namespace ‘foo’, arg should be fN</p>
</dd>
<dt>affix <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">generate prefixes/suffixes of features; argument ‘+2a,-3b,+1’ means generate 2-char prefixes for
namespace a, 3-char suffixes for b and 1 char prefixes for default namespace</p>
</dd>
<dt>spelling <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">compute spelling features for a give namespace (use ‘_’ for default namespace)</p>
</dd>
<dt>dictionary <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">read a dictionary for additional features (arg either ‘x:file’ or just ‘file’)</p>
</dd>
<dt>dictionary_path <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">look in this directory for dictionaries; defaults to current directory or env{PATH}</p>
</dd>
<dt>interactions <span class="classifier-delimiter">:</span> <span class="classifier">str</span></dt>
<dd><p class="first last">Create feature interactions of any level between namespaces.</p>
</dd>
<dt>permutations <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Use permutations instead of combinations for feature interactions of same namespace.</p>
</dd>
<dt>leave_duplicate_interactions <span class="classifier-delimiter">:</span> <span class="classifier">bool</span></dt>
<dd><p class="first last">Don’t remove interactions with duplicate combinations of namespaces. For