forked from apachecn/pandas-doc-zh
-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathreshaping.html
1050 lines (975 loc) · 116 KB
/
reshaping.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
<span id="reshaping"></span><h1><span class="yiyi-st" id="yiyi-62">Reshaping and Pivot Tables</span></h1>
<blockquote>
<p>原文：<a href="http://pandas.pydata.org/pandas-docs/stable/reshaping.html">http://pandas.pydata.org/pandas-docs/stable/reshaping.html</a></p>
<p>译者：<a href="https://github.com/wizardforcel">飞龙</a> <a href="http://usyiyi.cn/">UsyiyiCN</a></p>
<p>校对：（虚位以待）</p>
</blockquote>
<div class="section" id="reshaping-by-pivoting-dataframe-objects">
<h2><span class="yiyi-st" id="yiyi-63">Reshaping by pivoting DataFrame objects</span></h2>
<p><span class="yiyi-st" id="yiyi-64">数据通常以所谓的“堆叠”或“记录”格式存储在CSV文件或数据库中：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [1]: </span><span class="n">df</span>
<span class="gr">Out[1]: </span>
<span class="go"> date variable value</span>
<span class="go">0 2000-01-03 A 0.469112</span>
<span class="go">1 2000-01-04 A -0.282863</span>
<span class="go">2 2000-01-05 A -1.509059</span>
<span class="go">3 2000-01-03 B -1.135632</span>
<span class="go">4 2000-01-04 B 1.212112</span>
<span class="go">5 2000-01-05 B -0.173215</span>
<span class="go">6 2000-01-03 C 0.119209</span>
<span class="go">7 2000-01-04 C -1.044236</span>
<span class="go">8 2000-01-05 C -0.861849</span>
<span class="go">9 2000-01-03 D -2.104569</span>
<span class="go">10 2000-01-04 D -0.494929</span>
<span class="go">11 2000-01-05 D 1.071804</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-65">对于好奇这里是如何创建上面的DataFrame：</span></p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">pandas.util.testing</span> <span class="kn">as</span> <span class="nn">tm</span><span class="p">;</span> <span class="n">tm</span><span class="o">.</span><span class="n">N</span> <span class="o">=</span> <span class="mi">3</span>
<span class="k">def</span> <span class="nf">unpivot</span><span class="p">(</span><span class="n">frame</span><span class="p">):</span>
<span class="n">N</span><span class="p">,</span> <span class="n">K</span> <span class="o">=</span> <span class="n">frame</span><span class="o">.</span><span class="n">shape</span>
<span class="n">data</span> <span class="o">=</span> <span class="p">{</span><span class="s1">'value'</span> <span class="p">:</span> <span class="n">frame</span><span class="o">.</span><span class="n">values</span><span class="o">.</span><span class="n">ravel</span><span class="p">(</span><span class="s1">'F'</span><span class="p">),</span>
<span class="s1">'variable'</span> <span class="p">:</span> <span class="n">np</span><span class="o">.</span><span class="n">asarray</span><span class="p">(</span><span class="n">frame</span><span class="o">.</span><span class="n">columns</span><span class="p">)</span><span class="o">.</span><span class="n">repeat</span><span class="p">(</span><span class="n">N</span><span class="p">),</span>
<span class="s1">'date'</span> <span class="p">:</span> <span class="n">np</span><span class="o">.</span><span class="n">tile</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">asarray</span><span class="p">(</span><span class="n">frame</span><span class="o">.</span><span class="n">index</span><span class="p">),</span> <span class="n">K</span><span class="p">)}</span>
<span class="k">return</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span><span class="n">data</span><span class="p">,</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'date'</span><span class="p">,</span> <span class="s1">'variable'</span><span class="p">,</span> <span class="s1">'value'</span><span class="p">])</span>
<span class="n">df</span> <span class="o">=</span> <span class="n">unpivot</span><span class="p">(</span><span class="n">tm</span><span class="o">.</span><span class="n">makeTimeDataFrame</span><span class="p">())</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-66">要选择变量<code class="docutils literal"><span class="pre">A</span></code>的所有内容，我们可以：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [2]: </span><span class="n">df</span><span class="p">[</span><span class="n">df</span><span class="p">[</span><span class="s1">'variable'</span><span class="p">]</span> <span class="o">==</span> <span class="s1">'A'</span><span class="p">]</span>
<span class="gr">Out[2]: </span>
<span class="go"> date variable value</span>
<span class="go">0 2000-01-03 A 0.469112</span>
<span class="go">1 2000-01-04 A -0.282863</span>
<span class="go">2 2000-01-05 A -1.509059</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-67">但是假设我们希望用变量进行时间序列运算。</span><span class="yiyi-st" id="yiyi-68">更好的表示是<code class="docutils literal"><span class="pre">columns</span></code>是唯一变量，<code class="docutils literal"><span class="pre">index</span></code>标识个别观察。</span><span class="yiyi-st" id="yiyi-69">要将数据重新整形为此表单，请使用<code class="docutils literal"><span class="pre">pivot</span></code>函数：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [3]: </span><span class="n">df</span><span class="o">.</span><span class="n">pivot</span><span class="p">(</span><span class="n">index</span><span class="o">=</span><span class="s1">'date'</span><span class="p">,</span> <span class="n">columns</span><span class="o">=</span><span class="s1">'variable'</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="s1">'value'</span><span class="p">)</span>
<span class="gr">Out[3]: </span>
<span class="go">variable A B C D</span>
<span class="go">date </span>
<span class="go">2000-01-03 0.469112 -1.135632 0.119209 -2.104569</span>
<span class="go">2000-01-04 -0.282863 1.212112 -1.044236 -0.494929</span>
<span class="go">2000-01-05 -1.509059 -0.173215 -0.861849 1.071804</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-70">如果省略<code class="docutils literal"><span class="pre">values</span></code>参数，并且输入DataFrame具有多个不用作<code class="docutils literal"><span class="pre">pivot</span></code>的列或索引输入的值列，则生成的“pivoted” DataFrame将具有<a class="reference internal" href="advanced.html#advanced-hierarchical"><span class="std std-ref">hierarchical columns</span></a>，其最高级别表示相应的值列：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [4]: </span><span class="n">df</span><span class="p">[</span><span class="s1">'value2'</span><span class="p">]</span> <span class="o">=</span> <span class="n">df</span><span class="p">[</span><span class="s1">'value'</span><span class="p">]</span> <span class="o">*</span> <span class="mi">2</span>
<span class="gp">In [5]: </span><span class="n">pivoted</span> <span class="o">=</span> <span class="n">df</span><span class="o">.</span><span class="n">pivot</span><span class="p">(</span><span class="s1">'date'</span><span class="p">,</span> <span class="s1">'variable'</span><span class="p">)</span>
<span class="gp">In [6]: </span><span class="n">pivoted</span>
<span class="gr">Out[6]: </span>
<span class="go"> value value2 \</span>
<span class="go">variable A B C D A B </span>
<span class="go">date </span>
<span class="go">2000-01-03 0.469112 -1.135632 0.119209 -2.104569 0.938225 -2.271265 </span>
<span class="go">2000-01-04 -0.282863 1.212112 -1.044236 -0.494929 -0.565727 2.424224 </span>
<span class="go">2000-01-05 -1.509059 -0.173215 -0.861849 1.071804 -3.018117 -0.346429 </span>
<span class="go"> </span>
<span class="go">variable C D </span>
<span class="go">date </span>
<span class="go">2000-01-03 0.238417 -4.209138 </span>
<span class="go">2000-01-04 -2.088472 -0.989859 </span>
<span class="go">2000-01-05 -1.723698 2.143608 </span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-71">您当然可以从透视的DataFrame中选择子集：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [7]: </span><span class="n">pivoted</span><span class="p">[</span><span class="s1">'value2'</span><span class="p">]</span>
<span class="gr">Out[7]: </span>
<span class="go">variable A B C D</span>
<span class="go">date </span>
<span class="go">2000-01-03 0.938225 -2.271265 0.238417 -4.209138</span>
<span class="go">2000-01-04 -0.565727 2.424224 -2.088472 -0.989859</span>
<span class="go">2000-01-05 -3.018117 -0.346429 -1.723698 2.143608</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-72">请注意，在数据是均匀类型的情况下，这将返回基础数据的视图。</span></p>
</div>
<div class="section" id="reshaping-by-stacking-and-unstacking">
<span id="reshaping-stacking"></span><h2><span class="yiyi-st" id="yiyi-73">Reshaping by stacking and unstacking</span></h2>
<p><span class="yiyi-st" id="yiyi-74">与<code class="docutils literal"><span class="pre">pivot</span></code>函数紧密相关的是当前在Series和DataFrame上可用的相关<code class="docutils literal"><span class="pre">stack</span></code>和<code class="docutils literal"><span class="pre">unstack</span></code>功能。</span><span class="yiyi-st" id="yiyi-75">这些函数设计为与<code class="docutils literal"><span class="pre">MultiIndex</span></code>对象一起使用（请参阅<a class="reference internal" href="advanced.html#advanced-hierarchical"><span class="std std-ref">hierarchical indexing</span></a>一节）。</span><span class="yiyi-st" id="yiyi-76">这些功能基本上是：</span></p>
<blockquote>
<div><ul class="simple">
<li><span class="yiyi-st" id="yiyi-77"><code class="docutils literal"><span class="pre">stack</span></code>：“枢轴”（可能是层次结构）列标签的级别，返回具有新的最内层行标签的索引的DataFrame。</span></li>
<li><span class="yiyi-st" id="yiyi-78"><code class="docutils literal"><span class="pre">unstack</span></code>：从<code class="docutils literal"><span class="pre">stack</span></code>的反向操作：将（可能是分层的）行索引的级别“枢转”到列轴，产生具有新的最内层级的重新形成的DataFrame的列标签。</span></li>
</ul>
</div></blockquote>
<p><span class="yiyi-st" id="yiyi-79">最清楚的解释是通过例子。</span><span class="yiyi-st" id="yiyi-80">让我们从分层索引部分获取前面的示例数据集：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [8]: </span><span class="n">tuples</span> <span class="o">=</span> <span class="nb">list</span><span class="p">(</span><span class="nb">zip</span><span class="p">(</span><span class="o">*</span><span class="p">[[</span><span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'baz'</span><span class="p">,</span> <span class="s1">'baz'</span><span class="p">,</span>
<span class="gp"> ...:</span> <span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'qux'</span><span class="p">,</span> <span class="s1">'qux'</span><span class="p">],</span>
<span class="gp"> ...:</span> <span class="p">[</span><span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span><span class="p">,</span> <span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span><span class="p">,</span>
<span class="gp"> ...:</span> <span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span><span class="p">,</span> <span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span><span class="p">]]))</span>
<span class="gp"> ...:</span>
<span class="gp">In [9]: </span><span class="n">index</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">MultiIndex</span><span class="o">.</span><span class="n">from_tuples</span><span class="p">(</span><span class="n">tuples</span><span class="p">,</span> <span class="n">names</span><span class="o">=</span><span class="p">[</span><span class="s1">'first'</span><span class="p">,</span> <span class="s1">'second'</span><span class="p">])</span>
<span class="gp">In [10]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">8</span><span class="p">,</span> <span class="mi">2</span><span class="p">),</span> <span class="n">index</span><span class="o">=</span><span class="n">index</span><span class="p">,</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">])</span>
<span class="gp">In [11]: </span><span class="n">df2</span> <span class="o">=</span> <span class="n">df</span><span class="p">[:</span><span class="mi">4</span><span class="p">]</span>
<span class="gp">In [12]: </span><span class="n">df2</span>
<span class="gr">Out[12]: </span>
<span class="go"> A B</span>
<span class="go">first second </span>
<span class="go">bar one 0.721555 -0.706771</span>
<span class="go"> two -1.039575 0.271860</span>
<span class="go">baz one -0.424972 0.567020</span>
<span class="go"> two 0.276232 -1.087401</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-81"><code class="docutils literal"><span class="pre">stack</span></code>函数“压缩”DataFrame的列中的一个级别，以产生：</span></p>
<blockquote>
<div><ul class="simple">
<li><span class="yiyi-st" id="yiyi-82">A系列，在一个简单的列索引的情况下</span></li>
<li><span class="yiyi-st" id="yiyi-83">在列中的<code class="docutils literal"><span class="pre">MultiIndex</span></code>的情况下为DataFrame</span></li>
</ul>
</div></blockquote>
<p><span class="yiyi-st" id="yiyi-84">如果列具有<code class="docutils literal"><span class="pre">MultiIndex</span></code>，您可以选择要堆叠的级别。</span><span class="yiyi-st" id="yiyi-85">堆叠级别成为列上的<code class="docutils literal"><span class="pre">MultiIndex</span></code>中的新的最低级别：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [13]: </span><span class="n">stacked</span> <span class="o">=</span> <span class="n">df2</span><span class="o">.</span><span class="n">stack</span><span class="p">()</span>
<span class="gp">In [14]: </span><span class="n">stacked</span>
<span class="gr">Out[14]: </span>
<span class="go">first second </span>
<span class="go">bar one A 0.721555</span>
<span class="go"> B -0.706771</span>
<span class="go"> two A -1.039575</span>
<span class="go"> B 0.271860</span>
<span class="go">baz one A -0.424972</span>
<span class="go"> B 0.567020</span>
<span class="go"> two A 0.276232</span>
<span class="go"> B -1.087401</span>
<span class="go">dtype: float64</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-86">With a “stacked” DataFrame or Series (having a <code class="docutils literal"><span class="pre">MultiIndex</span></code> as the <code class="docutils literal"><span class="pre">index</span></code>), the inverse operation of <code class="docutils literal"><span class="pre">stack</span></code> is <code class="docutils literal"><span class="pre">unstack</span></code>, which by default unstacks the <strong>last level</strong>:</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [15]: </span><span class="n">stacked</span><span class="o">.</span><span class="n">unstack</span><span class="p">()</span>
<span class="gr">Out[15]: </span>
<span class="go"> A B</span>
<span class="go">first second </span>
<span class="go">bar one 0.721555 -0.706771</span>
<span class="go"> two -1.039575 0.271860</span>
<span class="go">baz one -0.424972 0.567020</span>
<span class="go"> two 0.276232 -1.087401</span>
<span class="gp">In [16]: </span><span class="n">stacked</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="gr">Out[16]: </span>
<span class="go">second one two</span>
<span class="go">first </span>
<span class="go">bar A 0.721555 -1.039575</span>
<span class="go"> B -0.706771 0.271860</span>
<span class="go">baz A -0.424972 0.276232</span>
<span class="go"> B 0.567020 -1.087401</span>
<span class="gp">In [17]: </span><span class="n">stacked</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="mi">0</span><span class="p">)</span>
<span class="gr">Out[17]: </span>
<span class="go">first bar baz</span>
<span class="go">second </span>
<span class="go">one A 0.721555 -0.424972</span>
<span class="go"> B -0.706771 0.567020</span>
<span class="go">two A -1.039575 0.276232</span>
<span class="go"> B 0.271860 -1.087401</span>
</pre></div>
</div>
<p id="reshaping-unstack-by-name"><span class="yiyi-st" id="yiyi-87">如果索引具有名称，则可以使用级别名称，而不是指定级别号：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [18]: </span><span class="n">stacked</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="s1">'second'</span><span class="p">)</span>
<span class="gr">Out[18]: </span>
<span class="go">second one two</span>
<span class="go">first </span>
<span class="go">bar A 0.721555 -1.039575</span>
<span class="go"> B -0.706771 0.271860</span>
<span class="go">baz A -0.424972 0.276232</span>
<span class="go"> B 0.567020 -1.087401</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-88">请注意，<code class="docutils literal"><span class="pre">stack</span></code>和<code class="docutils literal"><span class="pre">unstack</span></code>方法隐含地排序所涉及的索引级别。</span><span class="yiyi-st" id="yiyi-89">因此，调用<code class="docutils literal"><span class="pre">stack</span></code>然后<code class="docutils literal"><span class="pre">unstack</span></code>，反之亦然，将导致原始DataFrame或系列的<strong>排序</strong>副本：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [19]: </span><span class="n">index</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">MultiIndex</span><span class="o">.</span><span class="n">from_product</span><span class="p">([[</span><span class="mi">2</span><span class="p">,</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="s1">'a'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">]])</span>
<span class="gp">In [20]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">4</span><span class="p">),</span> <span class="n">index</span><span class="o">=</span><span class="n">index</span><span class="p">,</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">])</span>
<span class="gp">In [21]: </span><span class="n">df</span>
<span class="gr">Out[21]: </span>
<span class="go"> A</span>
<span class="go">2 a -0.370647</span>
<span class="go"> b -1.157892</span>
<span class="go">1 a -1.344312</span>
<span class="go"> b 0.844885</span>
<span class="gp">In [22]: </span><span class="nb">all</span><span class="p">(</span><span class="n">df</span><span class="o">.</span><span class="n">unstack</span><span class="p">()</span><span class="o">.</span><span class="n">stack</span><span class="p">()</span> <span class="o">==</span> <span class="n">df</span><span class="o">.</span><span class="n">sort_index</span><span class="p">())</span>
<span class="gr">Out[22]: </span><span class="bp">True</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-90">而如果去除对<code class="docutils literal"><span class="pre">sort_index</span></code>的调用，则上述代码将产生<code class="docutils literal"><span class="pre">TypeError</span></code>。</span></p>
<div class="section" id="multiple-levels">
<span id="reshaping-stack-multiple"></span><h3><span class="yiyi-st" id="yiyi-91">Multiple Levels</span></h3>
<p><span class="yiyi-st" id="yiyi-92">您还可以通过传递级别列表来一次堆叠或取消堆栈多个级别，在这种情况下，最终结果就好像单独处理列表中的每个级别一样。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [23]: </span><span class="n">columns</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">MultiIndex</span><span class="o">.</span><span class="n">from_tuples</span><span class="p">([</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'cat'</span><span class="p">,</span> <span class="s1">'long'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'B'</span><span class="p">,</span> <span class="s1">'cat'</span><span class="p">,</span> <span class="s1">'long'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'dog'</span><span class="p">,</span> <span class="s1">'short'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'B'</span><span class="p">,</span> <span class="s1">'dog'</span><span class="p">,</span> <span class="s1">'short'</span><span class="p">)</span>
<span class="gp"> ....:</span> <span class="p">],</span>
<span class="gp"> ....:</span> <span class="n">names</span><span class="o">=</span><span class="p">[</span><span class="s1">'exp'</span><span class="p">,</span> <span class="s1">'animal'</span><span class="p">,</span> <span class="s1">'hair_length'</span><span class="p">]</span>
<span class="gp"> ....:</span> <span class="p">)</span>
<span class="gp"> ....:</span>
<span class="gp">In [24]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">),</span> <span class="n">columns</span><span class="o">=</span><span class="n">columns</span><span class="p">)</span>
<span class="gp">In [25]: </span><span class="n">df</span>
<span class="gr">Out[25]: </span>
<span class="go">exp A B A B</span>
<span class="go">animal cat cat dog dog</span>
<span class="go">hair_length long long short short</span>
<span class="go">0 1.075770 -0.109050 1.643563 -1.469388</span>
<span class="go">1 0.357021 -0.674600 -1.776904 -0.968914</span>
<span class="go">2 -1.294524 0.413738 0.276662 -0.472035</span>
<span class="go">3 -0.013960 -0.362543 -0.006154 -0.923061</span>
<span class="gp">In [26]: </span><span class="n">df</span><span class="o">.</span><span class="n">stack</span><span class="p">(</span><span class="n">level</span><span class="o">=</span><span class="p">[</span><span class="s1">'animal'</span><span class="p">,</span> <span class="s1">'hair_length'</span><span class="p">])</span>
<span class="gr">Out[26]: </span>
<span class="go">exp A B</span>
<span class="go"> animal hair_length </span>
<span class="go">0 cat long 1.075770 -0.109050</span>
<span class="go"> dog short 1.643563 -1.469388</span>
<span class="go">1 cat long 0.357021 -0.674600</span>
<span class="go"> dog short -1.776904 -0.968914</span>
<span class="go">2 cat long -1.294524 0.413738</span>
<span class="go"> dog short 0.276662 -0.472035</span>
<span class="go">3 cat long -0.013960 -0.362543</span>
<span class="go"> dog short -0.006154 -0.923061</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-93">级别列表可以包含级别名称或级别号（但不能是两者的混合）。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="c"># df.stack(level=['animal', 'hair_length'])</span>
<span class="c"># from above is equivalent to:</span>
<span class="gp">In [27]: </span><span class="n">df</span><span class="o">.</span><span class="n">stack</span><span class="p">(</span><span class="n">level</span><span class="o">=</span><span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">])</span>
<span class="gr">Out[27]: </span>
<span class="go">exp A B</span>
<span class="go"> animal hair_length </span>
<span class="go">0 cat long 1.075770 -0.109050</span>
<span class="go"> dog short 1.643563 -1.469388</span>
<span class="go">1 cat long 0.357021 -0.674600</span>
<span class="go"> dog short -1.776904 -0.968914</span>
<span class="go">2 cat long -1.294524 0.413738</span>
<span class="go"> dog short 0.276662 -0.472035</span>
<span class="go">3 cat long -0.013960 -0.362543</span>
<span class="go"> dog short -0.006154 -0.923061</span>
</pre></div>
</div>
</div>
<div class="section" id="missing-data">
<h3><span class="yiyi-st" id="yiyi-94">Missing Data</span></h3>
<p><span class="yiyi-st" id="yiyi-95">这些函数在处理缺失数据方面是智能的，并且不期望分层索引中的每个子组具有相同的一组标签。</span><span class="yiyi-st" id="yiyi-96">它们还可以处理未排序的索引（但是您可以通过调用<code class="docutils literal"><span class="pre">sort_index</span></code>来排序）。</span><span class="yiyi-st" id="yiyi-97">这里有一个更复杂的例子：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [28]: </span><span class="n">columns</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">MultiIndex</span><span class="o">.</span><span class="n">from_tuples</span><span class="p">([(</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'cat'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'B'</span><span class="p">,</span> <span class="s1">'dog'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'B'</span><span class="p">,</span> <span class="s1">'cat'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'dog'</span><span class="p">)],</span>
<span class="gp"> ....:</span> <span class="n">names</span><span class="o">=</span><span class="p">[</span><span class="s1">'exp'</span><span class="p">,</span> <span class="s1">'animal'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [29]: </span><span class="n">index</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">MultiIndex</span><span class="o">.</span><span class="n">from_product</span><span class="p">([(</span><span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'baz'</span><span class="p">,</span> <span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'qux'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span><span class="p">)],</span>
<span class="gp"> ....:</span> <span class="n">names</span><span class="o">=</span><span class="p">[</span><span class="s1">'first'</span><span class="p">,</span> <span class="s1">'second'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [30]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">8</span><span class="p">,</span> <span class="mi">4</span><span class="p">),</span> <span class="n">index</span><span class="o">=</span><span class="n">index</span><span class="p">,</span> <span class="n">columns</span><span class="o">=</span><span class="n">columns</span><span class="p">)</span>
<span class="gp">In [31]: </span><span class="n">df2</span> <span class="o">=</span> <span class="n">df</span><span class="o">.</span><span class="n">ix</span><span class="p">[[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">5</span><span class="p">,</span> <span class="mi">7</span><span class="p">]]</span>
<span class="gp">In [32]: </span><span class="n">df2</span>
<span class="gr">Out[32]: </span>
<span class="go">exp A B A</span>
<span class="go">animal cat dog cat dog</span>
<span class="go">first second </span>
<span class="go">bar one 0.895717 0.805244 -1.206412 2.565646</span>
<span class="go"> two 1.431256 1.340309 -1.170299 -0.226169</span>
<span class="go">baz one 0.410835 0.813850 0.132003 -0.827317</span>
<span class="go">foo one -1.413681 1.607920 1.024180 0.569605</span>
<span class="go"> two 0.875906 -2.211372 0.974466 -2.006747</span>
<span class="go">qux two -1.226825 0.769804 -1.281247 -0.727707</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-98">如上所述，可以使用<code class="docutils literal"><span class="pre">level</span></code>参数调用<code class="docutils literal"><span class="pre">stack</span></code>，以选择要堆叠的列中的哪个级别：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [33]: </span><span class="n">df2</span><span class="o">.</span><span class="n">stack</span><span class="p">(</span><span class="s1">'exp'</span><span class="p">)</span>
<span class="gr">Out[33]: </span>
<span class="go">animal cat dog</span>
<span class="go">first second exp </span>
<span class="go">bar one A 0.895717 2.565646</span>
<span class="go"> B -1.206412 0.805244</span>
<span class="go"> two A 1.431256 -0.226169</span>
<span class="go"> B -1.170299 1.340309</span>
<span class="go">baz one A 0.410835 -0.827317</span>
<span class="go"> B 0.132003 0.813850</span>
<span class="go">foo one A -1.413681 0.569605</span>
<span class="go"> B 1.024180 1.607920</span>
<span class="go"> two A 0.875906 -2.006747</span>
<span class="go"> B 0.974466 -2.211372</span>
<span class="go">qux two A -1.226825 -0.727707</span>
<span class="go"> B -1.281247 0.769804</span>
<span class="gp">In [34]: </span><span class="n">df2</span><span class="o">.</span><span class="n">stack</span><span class="p">(</span><span class="s1">'animal'</span><span class="p">)</span>
<span class="gr">Out[34]: </span>
<span class="go">exp A B</span>
<span class="go">first second animal </span>
<span class="go">bar one cat 0.895717 -1.206412</span>
<span class="go"> dog 2.565646 0.805244</span>
<span class="go"> two cat 1.431256 -1.170299</span>
<span class="go"> dog -0.226169 1.340309</span>
<span class="go">baz one cat 0.410835 0.132003</span>
<span class="go"> dog -0.827317 0.813850</span>
<span class="go">foo one cat -1.413681 1.024180</span>
<span class="go"> dog 0.569605 1.607920</span>
<span class="go"> two cat 0.875906 0.974466</span>
<span class="go"> dog -2.006747 -2.211372</span>
<span class="go">qux two cat -1.226825 -1.281247</span>
<span class="go"> dog -0.727707 0.769804</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-99">如果子组不具有相同的标签集，则解除堆叠可能导致缺失值。</span><span class="yiyi-st" id="yiyi-100">默认情况下，缺少的值将替换为该数据类型的默认填充值，<code class="docutils literal"><span class="pre">NaN</span></code>用于float，<code class="docutils literal"><span class="pre">NaT</span></code>用于datetimelike等。</span><span class="yiyi-st" id="yiyi-101">对于整数类型，默认情况下，数据将转换为float，缺少的值将设置为<code class="docutils literal"><span class="pre">NaN</span></code>。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [35]: </span><span class="n">df3</span> <span class="o">=</span> <span class="n">df</span><span class="o">.</span><span class="n">iloc</span><span class="p">[[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">7</span><span class="p">],</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">]]</span>
<span class="gp">In [36]: </span><span class="n">df3</span>
<span class="gr">Out[36]: </span>
<span class="go">exp B </span>
<span class="go">animal dog cat</span>
<span class="go">first second </span>
<span class="go">bar one 0.805244 -1.206412</span>
<span class="go"> two 1.340309 -1.170299</span>
<span class="go">foo one 1.607920 1.024180</span>
<span class="go">qux two 0.769804 -1.281247</span>
<span class="gp">In [37]: </span><span class="n">df3</span><span class="o">.</span><span class="n">unstack</span><span class="p">()</span>
<span class="gr">Out[37]: </span>
<span class="go">exp B </span>
<span class="go">animal dog cat </span>
<span class="go">second one two one two</span>
<span class="go">first </span>
<span class="go">bar 0.805244 1.340309 -1.206412 -1.170299</span>
<span class="go">foo 1.607920 NaN 1.024180 NaN</span>
<span class="go">qux NaN 0.769804 NaN -1.281247</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-102">或者，unstack采用可选的<code class="docutils literal"><span class="pre">fill_value</span></code>参数，用于指定缺少的数据的值。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [38]: </span><span class="n">df3</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="n">fill_value</span><span class="o">=-</span><span class="mf">1e9</span><span class="p">)</span>
<span class="gr">Out[38]: </span>
<span class="go">exp B </span>
<span class="go">animal dog cat </span>
<span class="go">second one two one two</span>
<span class="go">first </span>
<span class="go">bar 8.052440e-01 1.340309e+00 -1.206412e+00 -1.170299e+00</span>
<span class="go">foo 1.607920e+00 -1.000000e+09 1.024180e+00 -1.000000e+09</span>
<span class="go">qux -1.000000e+09 7.698036e-01 -1.000000e+09 -1.281247e+00</span>
</pre></div>
</div>
</div>
<div class="section" id="with-a-multiindex">
<h3><span class="yiyi-st" id="yiyi-103">With a MultiIndex</span></h3>
<p><span class="yiyi-st" id="yiyi-104">当列为<code class="docutils literal"><span class="pre">MultiIndex</span></code>时解除堆叠也小心做正确的事：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [39]: </span><span class="n">df</span><span class="p">[:</span><span class="mi">3</span><span class="p">]</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="mi">0</span><span class="p">)</span>
<span class="gr">Out[39]: </span>
<span class="go">exp A B A \</span>
<span class="go">animal cat dog cat dog </span>
<span class="go">first bar baz bar baz bar baz bar </span>
<span class="go">second </span>
<span class="go">one 0.895717 0.410835 0.805244 0.81385 -1.206412 0.132003 2.565646 </span>
<span class="go">two 1.431256 NaN 1.340309 NaN -1.170299 NaN -0.226169 </span>
<span class="go">exp </span>
<span class="go">animal </span>
<span class="go">first baz </span>
<span class="go">second </span>
<span class="go">one -0.827317 </span>
<span class="go">two NaN </span>
<span class="gp">In [40]: </span><span class="n">df2</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="gr">Out[40]: </span>
<span class="go">exp A B A \</span>
<span class="go">animal cat dog cat dog </span>
<span class="go">second one two one two one two one </span>
<span class="go">first </span>
<span class="go">bar 0.895717 1.431256 0.805244 1.340309 -1.206412 -1.170299 2.565646 </span>
<span class="go">baz 0.410835 NaN 0.813850 NaN 0.132003 NaN -0.827317 </span>
<span class="go">foo -1.413681 0.875906 1.607920 -2.211372 1.024180 0.974466 0.569605 </span>
<span class="go">qux NaN -1.226825 NaN 0.769804 NaN -1.281247 NaN </span>
<span class="go">exp </span>
<span class="go">animal </span>
<span class="go">second two </span>
<span class="go">first </span>
<span class="go">bar -0.226169 </span>
<span class="go">baz NaN </span>
<span class="go">foo -2.006747 </span>
<span class="go">qux -0.727707 </span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="reshaping-by-melt">
<span id="reshaping-melt"></span><h2><span class="yiyi-st" id="yiyi-105">Reshaping by Melt</span></h2>
<p><span class="yiyi-st" id="yiyi-106"><a class="reference internal" href="generated/pandas.melt.html#pandas.melt" title="pandas.melt"><code class="xref py py-func docutils literal"><span class="pre">melt()</span></code></a>函数有助于将DataFrame压缩为一个或多个列是标识符变量的格式，而所有其他列（被认为是测量变量）都与行轴“不相关”，只留下两个非标识符列，“variable”和“value”。</span><span class="yiyi-st" id="yiyi-107">可以通过提供<code class="docutils literal"><span class="pre">var_name</span></code>和<code class="docutils literal"><span class="pre">value_name</span></code>参数来自定义这些列的名称。</span></p>
<p><span class="yiyi-st" id="yiyi-108">例如，</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [41]: </span><span class="n">cheese</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'first'</span> <span class="p">:</span> <span class="p">[</span><span class="s1">'John'</span><span class="p">,</span> <span class="s1">'Mary'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'last'</span> <span class="p">:</span> <span class="p">[</span><span class="s1">'Doe'</span><span class="p">,</span> <span class="s1">'Bo'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'height'</span> <span class="p">:</span> <span class="p">[</span><span class="mf">5.5</span><span class="p">,</span> <span class="mf">6.0</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'weight'</span> <span class="p">:</span> <span class="p">[</span><span class="mi">130</span><span class="p">,</span> <span class="mi">150</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [42]: </span><span class="n">cheese</span>
<span class="gr">Out[42]: </span>
<span class="go"> first height last weight</span>
<span class="go">0 John 5.5 Doe 130</span>
<span class="go">1 Mary 6.0 Bo 150</span>
<span class="gp">In [43]: </span><span class="n">pd</span><span class="o">.</span><span class="n">melt</span><span class="p">(</span><span class="n">cheese</span><span class="p">,</span> <span class="n">id_vars</span><span class="o">=</span><span class="p">[</span><span class="s1">'first'</span><span class="p">,</span> <span class="s1">'last'</span><span class="p">])</span>
<span class="gr">Out[43]: </span>
<span class="go"> first last variable value</span>
<span class="go">0 John Doe height 5.5</span>
<span class="go">1 Mary Bo height 6.0</span>
<span class="go">2 John Doe weight 130.0</span>
<span class="go">3 Mary Bo weight 150.0</span>
<span class="gp">In [44]: </span><span class="n">pd</span><span class="o">.</span><span class="n">melt</span><span class="p">(</span><span class="n">cheese</span><span class="p">,</span> <span class="n">id_vars</span><span class="o">=</span><span class="p">[</span><span class="s1">'first'</span><span class="p">,</span> <span class="s1">'last'</span><span class="p">],</span> <span class="n">var_name</span><span class="o">=</span><span class="s1">'quantity'</span><span class="p">)</span>
<span class="gr">Out[44]: </span>
<span class="go"> first last quantity value</span>
<span class="go">0 John Doe height 5.5</span>
<span class="go">1 Mary Bo height 6.0</span>
<span class="go">2 John Doe weight 130.0</span>
<span class="go">3 Mary Bo weight 150.0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-109">另一种变换方式是使用<code class="docutils literal"><span class="pre">wide_to_long</span></code>面板数据便利功能。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [45]: </span><span class="n">dft</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s2">"A1970"</span> <span class="p">:</span> <span class="p">{</span><span class="mi">0</span> <span class="p">:</span> <span class="s2">"a"</span><span class="p">,</span> <span class="mi">1</span> <span class="p">:</span> <span class="s2">"b"</span><span class="p">,</span> <span class="mi">2</span> <span class="p">:</span> <span class="s2">"c"</span><span class="p">},</span>
<span class="gp"> ....:</span> <span class="s2">"A1980"</span> <span class="p">:</span> <span class="p">{</span><span class="mi">0</span> <span class="p">:</span> <span class="s2">"d"</span><span class="p">,</span> <span class="mi">1</span> <span class="p">:</span> <span class="s2">"e"</span><span class="p">,</span> <span class="mi">2</span> <span class="p">:</span> <span class="s2">"f"</span><span class="p">},</span>
<span class="gp"> ....:</span> <span class="s2">"B1970"</span> <span class="p">:</span> <span class="p">{</span><span class="mi">0</span> <span class="p">:</span> <span class="mf">2.5</span><span class="p">,</span> <span class="mi">1</span> <span class="p">:</span> <span class="mf">1.2</span><span class="p">,</span> <span class="mi">2</span> <span class="p">:</span> <span class="o">.</span><span class="mi">7</span><span class="p">},</span>
<span class="gp"> ....:</span> <span class="s2">"B1980"</span> <span class="p">:</span> <span class="p">{</span><span class="mi">0</span> <span class="p">:</span> <span class="mf">3.2</span><span class="p">,</span> <span class="mi">1</span> <span class="p">:</span> <span class="mf">1.3</span><span class="p">,</span> <span class="mi">2</span> <span class="p">:</span> <span class="o">.</span><span class="mi">1</span><span class="p">},</span>
<span class="gp"> ....:</span> <span class="s2">"X"</span> <span class="p">:</span> <span class="nb">dict</span><span class="p">(</span><span class="nb">zip</span><span class="p">(</span><span class="nb">range</span><span class="p">(</span><span class="mi">3</span><span class="p">),</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">3</span><span class="p">)))</span>
<span class="gp"> ....:</span> <span class="p">})</span>
<span class="gp"> ....:</span>
<span class="gp">In [46]: </span><span class="n">dft</span><span class="p">[</span><span class="s2">"id"</span><span class="p">]</span> <span class="o">=</span> <span class="n">dft</span><span class="o">.</span><span class="n">index</span>
<span class="gp">In [47]: </span><span class="n">dft</span>
<span class="gr">Out[47]: </span>
<span class="go"> A1970 A1980 B1970 B1980 X id</span>
<span class="go">0 a d 2.5 3.2 -0.121306 0</span>
<span class="go">1 b e 1.2 1.3 -0.097883 1</span>
<span class="go">2 c f 0.7 0.1 0.695775 2</span>
<span class="gp">In [48]: </span><span class="n">pd</span><span class="o">.</span><span class="n">wide_to_long</span><span class="p">(</span><span class="n">dft</span><span class="p">,</span> <span class="p">[</span><span class="s2">"A"</span><span class="p">,</span> <span class="s2">"B"</span><span class="p">],</span> <span class="n">i</span><span class="o">=</span><span class="s2">"id"</span><span class="p">,</span> <span class="n">j</span><span class="o">=</span><span class="s2">"year"</span><span class="p">)</span>
<span class="gr">Out[48]: </span>
<span class="go"> X A B</span>
<span class="go">id year </span>
<span class="go">0 1970 -0.121306 a 2.5</span>
<span class="go">1 1970 -0.097883 b 1.2</span>
<span class="go">2 1970 0.695775 c 0.7</span>
<span class="go">0 1980 -0.121306 d 3.2</span>
<span class="go">1 1980 -0.097883 e 1.3</span>
<span class="go">2 1980 0.695775 f 0.1</span>
</pre></div>
</div>
</div>
<div class="section" id="combining-with-stats-and-groupby">
<h2><span class="yiyi-st" id="yiyi-110">Combining with stats and GroupBy</span></h2>
<p><span class="yiyi-st" id="yiyi-111">将<code class="docutils literal"><span class="pre">pivot</span></code> / <code class="docutils literal"><span class="pre">stack</span></code> / <code class="docutils literal"><span class="pre">unstack</span></code>与GroupBy和基本的Series和DataFrame统计函数组合可以产生一些非常有表现力和快速数据操作。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [49]: </span><span class="n">df</span>
<span class="gr">Out[49]: </span>
<span class="go">exp A B A</span>
<span class="go">animal cat dog cat dog</span>
<span class="go">first second </span>
<span class="go">bar one 0.895717 0.805244 -1.206412 2.565646</span>
<span class="go"> two 1.431256 1.340309 -1.170299 -0.226169</span>
<span class="go">baz one 0.410835 0.813850 0.132003 -0.827317</span>
<span class="go"> two -0.076467 -1.187678 1.130127 -1.436737</span>
<span class="go">foo one -1.413681 1.607920 1.024180 0.569605</span>
<span class="go"> two 0.875906 -2.211372 0.974466 -2.006747</span>
<span class="go">qux one -0.410001 -0.078638 0.545952 -1.219217</span>
<span class="go"> two -1.226825 0.769804 -1.281247 -0.727707</span>
<span class="gp">In [50]: </span><span class="n">df</span><span class="o">.</span><span class="n">stack</span><span class="p">()</span><span class="o">.</span><span class="n">mean</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span><span class="o">.</span><span class="n">unstack</span><span class="p">()</span>
<span class="gr">Out[50]: </span>
<span class="go">animal cat dog</span>
<span class="go">first second </span>
<span class="go">bar one -0.155347 1.685445</span>
<span class="go"> two 0.130479 0.557070</span>
<span class="go">baz one 0.271419 -0.006733</span>
<span class="go"> two 0.526830 -1.312207</span>
<span class="go">foo one -0.194750 1.088763</span>
<span class="go"> two 0.925186 -2.109060</span>
<span class="go">qux one 0.067976 -0.648927</span>
<span class="go"> two -1.254036 0.021048</span>
<span class="c"># same result, another way</span>
<span class="gp">In [51]: </span><span class="n">df</span><span class="o">.</span><span class="n">groupby</span><span class="p">(</span><span class="n">level</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">axis</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">()</span>
<span class="gr">Out[51]: </span>
<span class="go">animal cat dog</span>
<span class="go">first second </span>
<span class="go">bar one -0.155347 1.685445</span>
<span class="go"> two 0.130479 0.557070</span>
<span class="go">baz one 0.271419 -0.006733</span>
<span class="go"> two 0.526830 -1.312207</span>
<span class="go">foo one -0.194750 1.088763</span>
<span class="go"> two 0.925186 -2.109060</span>
<span class="go">qux one 0.067976 -0.648927</span>
<span class="go"> two -1.254036 0.021048</span>
<span class="gp">In [52]: </span><span class="n">df</span><span class="o">.</span><span class="n">stack</span><span class="p">()</span><span class="o">.</span><span class="n">groupby</span><span class="p">(</span><span class="n">level</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">()</span>
<span class="gr">Out[52]: </span>
<span class="go">exp A B</span>
<span class="go">second </span>
<span class="go">one 0.071448 0.455513</span>
<span class="go">two -0.424186 -0.204486</span>
<span class="gp">In [53]: </span><span class="n">df</span><span class="o">.</span><span class="n">mean</span><span class="p">()</span><span class="o">.</span><span class="n">unstack</span><span class="p">(</span><span class="mi">0</span><span class="p">)</span>
<span class="gr">Out[53]: </span>
<span class="go">exp A B</span>
<span class="go">animal </span>
<span class="go">cat 0.060843 0.018596</span>
<span class="go">dog -0.413580 0.232430</span>
</pre></div>
</div>
</div>
<div class="section" id="pivot-tables">
<h2><span class="yiyi-st" id="yiyi-112">Pivot tables</span></h2>
<p id="reshaping-pivot"><span class="yiyi-st" id="yiyi-113">函数<code class="docutils literal"><span class="pre">pandas.pivot_table</span></code>可用于创建电子表格样式的枢轴表。</span><span class="yiyi-st" id="yiyi-114">有关某些高级策略，请参阅<a class="reference internal" href="cookbook.html#cookbook-pivot"><span class="std std-ref">cookbook</span></a></span></p>
<p><span class="yiyi-st" id="yiyi-115">它需要一些参数</span></p>
<ul class="simple">
<li><span class="yiyi-st" id="yiyi-116"><code class="docutils literal"><span class="pre">data</span></code>：DataFrame对象</span></li>
<li><span class="yiyi-st" id="yiyi-117"><code class="docutils literal"><span class="pre">values</span></code>：要聚合的列或列的列表</span></li>
<li><span class="yiyi-st" id="yiyi-118"><code class="docutils literal"><span class="pre">index</span></code>：列，Grouper，与数据长度相同的数组或列表。</span><span class="yiyi-st" id="yiyi-119">按分组依据的数据透视表索引。</span><span class="yiyi-st" id="yiyi-120">如果传递数组，则其使用方式与列值相同。</span></li>
<li><span class="yiyi-st" id="yiyi-121"><code class="docutils literal"><span class="pre">columns</span></code>：列，Grouper，与数据长度相同的数组或列表。</span><span class="yiyi-st" id="yiyi-122">分组依据的关键字数据透视表列。</span><span class="yiyi-st" id="yiyi-123">如果传递数组，则其使用方式与列值相同。</span></li>
<li><span class="yiyi-st" id="yiyi-124"><code class="docutils literal"><span class="pre">aggfunc</span></code>：用于聚合的函数，默认为<code class="docutils literal"><span class="pre">numpy.mean</span></code></span></li>
</ul>
<p><span class="yiyi-st" id="yiyi-125">考虑一个像这样的数据集：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [54]: </span><span class="kn">import</span> <span class="nn">datetime</span>
<span class="gp">In [55]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'A'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'one'</span><span class="p">,</span> <span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span><span class="p">,</span> <span class="s1">'three'</span><span class="p">]</span> <span class="o">*</span> <span class="mi">6</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'B'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">,</span> <span class="s1">'C'</span><span class="p">]</span> <span class="o">*</span> <span class="mi">8</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'C'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'bar'</span><span class="p">]</span> <span class="o">*</span> <span class="mi">4</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">24</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="s1">'E'</span><span class="p">:</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">24</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="s1">'F'</span><span class="p">:</span> <span class="p">[</span><span class="n">datetime</span><span class="o">.</span><span class="n">datetime</span><span class="p">(</span><span class="mi">2013</span><span class="p">,</span> <span class="n">i</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">13</span><span class="p">)]</span> <span class="o">+</span>
<span class="gp"> ....:</span> <span class="p">[</span><span class="n">datetime</span><span class="o">.</span><span class="n">datetime</span><span class="p">(</span><span class="mi">2013</span><span class="p">,</span> <span class="n">i</span><span class="p">,</span> <span class="mi">15</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">13</span><span class="p">)]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [56]: </span><span class="n">df</span>
<span class="gr">Out[56]: </span>
<span class="go"> A B C D E F</span>
<span class="go">0 one A foo 0.341734 -0.317441 2013-01-01</span>
<span class="go">1 one B foo 0.959726 -1.236269 2013-02-01</span>
<span class="go">2 two C foo -1.110336 0.896171 2013-03-01</span>
<span class="go">3 three A bar -0.619976 -0.487602 2013-04-01</span>
<span class="go">4 one B bar 0.149748 -0.082240 2013-05-01</span>
<span class="go">5 one C bar -0.732339 -2.182937 2013-06-01</span>
<span class="go">6 two A foo 0.687738 0.380396 2013-07-01</span>
<span class="go">.. ... .. ... ... ... ...</span>
<span class="go">17 one C bar -0.345352 0.206053 2013-06-15</span>
<span class="go">18 two A foo 1.314232 -0.251905 2013-07-15</span>
<span class="go">19 three B foo 0.690579 -2.213588 2013-08-15</span>
<span class="go">20 one C foo 0.995761 1.063327 2013-09-15</span>
<span class="go">21 one A bar 2.396780 1.266143 2013-10-15</span>
<span class="go">22 two B bar 0.014871 0.299368 2013-11-15</span>
<span class="go">23 three C bar 3.357427 -0.863838 2013-12-15</span>
<span class="go">[24 rows x 6 columns]</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-126">我们可以非常容易地从这些数据生成数据透视表：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [57]: </span><span class="n">pd</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="s1">'D'</span><span class="p">,</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">],</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'C'</span><span class="p">])</span>
<span class="gr">Out[57]: </span>
<span class="go">C bar foo</span>
<span class="go">A B </span>
<span class="go">one A 1.120915 -0.514058</span>
<span class="go"> B -0.338421 0.002759</span>
<span class="go"> C -0.538846 0.699535</span>
<span class="go">three A -1.181568 NaN</span>
<span class="go"> B NaN 0.433512</span>
<span class="go"> C 0.588783 NaN</span>
<span class="go">two A NaN 1.000985</span>
<span class="go"> B 0.158248 NaN</span>
<span class="go"> C NaN 0.176180</span>
<span class="gp">In [58]: </span><span class="n">pd</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="s1">'D'</span><span class="p">,</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'B'</span><span class="p">],</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'C'</span><span class="p">],</span> <span class="n">aggfunc</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sum</span><span class="p">)</span>
<span class="gr">Out[58]: </span>
<span class="go">A one three two </span>
<span class="go">C bar foo bar foo bar foo</span>
<span class="go">B </span>
<span class="go">A 2.241830 -1.028115 -2.363137 NaN NaN 2.001971</span>
<span class="go">B -0.676843 0.005518 NaN 0.867024 0.316495 NaN</span>
<span class="go">C -1.077692 1.399070 1.177566 NaN NaN 0.352360</span>
<span class="gp">In [59]: </span><span class="n">pd</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="p">[</span><span class="s1">'D'</span><span class="p">,</span><span class="s1">'E'</span><span class="p">],</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'B'</span><span class="p">],</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'C'</span><span class="p">],</span> <span class="n">aggfunc</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sum</span><span class="p">)</span>
<span class="gr">Out[59]: </span>
<span class="go"> D E \</span>
<span class="go">A one three two one </span>
<span class="go">C bar foo bar foo bar foo bar </span>
<span class="go">B </span>
<span class="go">A 2.241830 -1.028115 -2.363137 NaN NaN 2.001971 2.786113 </span>
<span class="go">B -0.676843 0.005518 NaN 0.867024 0.316495 NaN 1.368280 </span>
<span class="go">C -1.077692 1.399070 1.177566 NaN NaN 0.352360 -1.976883 </span>
<span class="go"> </span>
<span class="go">A three two </span>
<span class="go">C foo bar foo bar foo </span>
<span class="go">B </span>
<span class="go">A -0.043211 1.922577 NaN NaN 0.128491 </span>
<span class="go">B -1.103384 NaN -2.128743 -0.194294 NaN </span>
<span class="go">C 1.495717 -0.263660 NaN NaN 0.872482 </span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-127">结果对象是在行和列上具有潜在分层索引的DataFrame。</span><span class="yiyi-st" id="yiyi-128">如果未给出<code class="docutils literal"><span class="pre">values</span></code>列名称，则数据透视表将包括可以在列中的附加层次级别中聚合的所有数据：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [60]: </span><span class="n">pd</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">],</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'C'</span><span class="p">])</span>
<span class="gr">Out[60]: </span>
<span class="go"> D E </span>
<span class="go">C bar foo bar foo</span>
<span class="go">A B </span>
<span class="go">one A 1.120915 -0.514058 1.393057 -0.021605</span>
<span class="go"> B -0.338421 0.002759 0.684140 -0.551692</span>
<span class="go"> C -0.538846 0.699535 -0.988442 0.747859</span>
<span class="go">three A -1.181568 NaN 0.961289 NaN</span>
<span class="go"> B NaN 0.433512 NaN -1.064372</span>
<span class="go"> C 0.588783 NaN -0.131830 NaN</span>
<span class="go">two A NaN 1.000985 NaN 0.064245</span>
<span class="go"> B 0.158248 NaN -0.097147 NaN</span>
<span class="go"> C NaN 0.176180 NaN 0.436241</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-129">此外，您还可以对<code class="docutils literal"><span class="pre">index</span></code>和<code class="docutils literal"><span class="pre">columns</span></code>关键字使用<code class="docutils literal"><span class="pre">Grouper</span></code>。</span><span class="yiyi-st" id="yiyi-130">有关<code class="docutils literal"><span class="pre">Grouper</span></code>的详细信息，请参阅<a class="reference internal" href="groupby.html#groupby-specify"><span class="std std-ref">Grouping with a Grouper specification</span></a>。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [61]: </span><span class="n">pd</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="s1">'D'</span><span class="p">,</span> <span class="n">index</span><span class="o">=</span><span class="n">pd</span><span class="o">.</span><span class="n">Grouper</span><span class="p">(</span><span class="n">freq</span><span class="o">=</span><span class="s1">'M'</span><span class="p">,</span> <span class="n">key</span><span class="o">=</span><span class="s1">'F'</span><span class="p">),</span> <span class="n">columns</span><span class="o">=</span><span class="s1">'C'</span><span class="p">)</span>
<span class="gr">Out[61]: </span>
<span class="go">C bar foo</span>
<span class="go">F </span>
<span class="go">2013-01-31 NaN -0.514058</span>
<span class="go">2013-02-28 NaN 0.002759</span>
<span class="go">2013-03-31 NaN 0.176180</span>
<span class="go">2013-04-30 -1.181568 NaN</span>
<span class="go">2013-05-31 -0.338421 NaN</span>
<span class="go">2013-06-30 -0.538846 NaN</span>
<span class="go">2013-07-31 NaN 1.000985</span>
<span class="go">2013-08-31 NaN 0.433512</span>
<span class="go">2013-09-30 NaN 0.699535</span>
<span class="go">2013-10-31 1.120915 NaN</span>
<span class="go">2013-11-30 0.158248 NaN</span>
<span class="go">2013-12-31 0.588783 NaN</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-131">如果您愿意，可以通过调用<code class="docutils literal"><span class="pre">to_string</span></code>来呈现表的一个不错的输出，省略缺少的值：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [62]: </span><span class="n">table</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">],</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'C'</span><span class="p">])</span>
<span class="gp">In [63]: </span><span class="k">print</span><span class="p">(</span><span class="n">table</span><span class="o">.</span><span class="n">to_string</span><span class="p">(</span><span class="n">na_rep</span><span class="o">=</span><span class="s1">''</span><span class="p">))</span>
<span class="go"> D E </span>
<span class="go">C bar foo bar foo</span>
<span class="go">A B </span>
<span class="go">one A 1.120915 -0.514058 1.393057 -0.021605</span>
<span class="go"> B -0.338421 0.002759 0.684140 -0.551692</span>
<span class="go"> C -0.538846 0.699535 -0.988442 0.747859</span>
<span class="go">three A -1.181568 0.961289 </span>
<span class="go"> B 0.433512 -1.064372</span>
<span class="go"> C 0.588783 -0.131830 </span>
<span class="go">two A 1.000985 0.064245</span>
<span class="go"> B 0.158248 -0.097147 </span>
<span class="go"> C 0.176180 0.436241</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-132">请注意，<code class="docutils literal"><span class="pre">pivot_table</span></code>也可用作DataFrame上的实例方法。</span></p>
<div class="section" id="adding-margins">
<span id="reshaping-pivot-margins"></span><h3><span class="yiyi-st" id="yiyi-133">Adding margins</span></h3>
<p><span class="yiyi-st" id="yiyi-134">如果您将<code class="docutils literal"><span class="pre">margins=True</span></code>传递到<code class="docutils literal"><span class="pre">pivot_table</span></code>，则特殊的<code class="docutils literal"><span class="pre">All</span></code>列和行将在行和列的类别中添加部分组聚合：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [64]: </span><span class="n">df</span><span class="o">.</span><span class="n">pivot_table</span><span class="p">(</span><span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">],</span> <span class="n">columns</span><span class="o">=</span><span class="s1">'C'</span><span class="p">,</span> <span class="n">margins</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span> <span class="n">aggfunc</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">std</span><span class="p">)</span>
<span class="gr">Out[64]: </span>
<span class="go"> D E </span>
<span class="go">C bar foo All bar foo All</span>
<span class="go">A B </span>
<span class="go">one A 1.804346 1.210272 1.569879 0.179483 0.418374 0.858005</span>
<span class="go"> B 0.690376 1.353355 0.898998 1.083825 0.968138 1.101401</span>
<span class="go"> C 0.273641 0.418926 0.771139 1.689271 0.446140 1.422136</span>
<span class="go">three A 0.794212 NaN 0.794212 2.049040 NaN 2.049040</span>
<span class="go"> B NaN 0.363548 0.363548 NaN 1.625237 1.625237</span>
<span class="go"> C 3.915454 NaN 3.915454 1.035215 NaN 1.035215</span>
<span class="go">two A NaN 0.442998 0.442998 NaN 0.447104 0.447104</span>
<span class="go"> B 0.202765 NaN 0.202765 0.560757 NaN 0.560757</span>
<span class="go"> C NaN 1.819408 1.819408 NaN 0.650439 0.650439</span>
<span class="go">All 1.556686 0.952552 1.246608 1.250924 0.899904 1.059389</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="cross-tabulations">
<span id="reshaping-crosstabulations"></span><h2><span class="yiyi-st" id="yiyi-135">Cross tabulations</span></h2>
<p><span class="yiyi-st" id="yiyi-136">使用<code class="docutils literal"><span class="pre">crosstab</span></code>函数计算两个（或多个）因子的交叉表。</span><span class="yiyi-st" id="yiyi-137">默认情况下，<code class="docutils literal"><span class="pre">crosstab</span></code>计算因子的频率表，除非传递值数组和聚合函数。</span></p>
<p><span class="yiyi-st" id="yiyi-138">它需要一些参数</span></p>
<ul class="simple">
<li><span class="yiyi-st" id="yiyi-139"><code class="docutils literal"><span class="pre">index</span></code>：array-like，在行中分组的值</span></li>
<li><span class="yiyi-st" id="yiyi-140"><code class="docutils literal"><span class="pre">columns</span></code>：array-like，在列中分组的值</span></li>
<li><span class="yiyi-st" id="yiyi-141"><code class="docutils literal"><span class="pre">values</span></code>：array-like，可选，根据因子聚合的值数组</span></li>
<li><span class="yiyi-st" id="yiyi-142"><code class="docutils literal"><span class="pre">aggfunc</span></code>：function，可选，如果未传递values数组，则计算频率表</span></li>
<li><span class="yiyi-st" id="yiyi-143"><code class="docutils literal"><span class="pre">rownames</span></code>：sequence，默认<code class="docutils literal"><span class="pre">None</span></code>必须匹配通过的行数组数</span></li>
<li><span class="yiyi-st" id="yiyi-144"><code class="docutils literal"><span class="pre">colnames</span></code>：序列，默认<code class="docutils literal"><span class="pre">None</span></code>（如果传递）必须匹配传递的列数组数</span></li>
<li><span class="yiyi-st" id="yiyi-145"><code class="docutils literal"><span class="pre">margins</span></code>：布尔值，默认值<code class="docutils literal"><span class="pre">False</span></code>，添加行/列边距（小计）</span></li>
<li><span class="yiyi-st" id="yiyi-146"><code class="docutils literal"><span class="pre">normalize</span></code>：boolean，{'all'，'index'，'columns'}或{0,1}，默认<code class="docutils literal"><span class="pre">False</span></code>。</span><span class="yiyi-st" id="yiyi-147">将所有值除以值的总和进行归一化。</span></li>
</ul>
<p><span class="yiyi-st" id="yiyi-148">任何传递的系列将使用其名称属性，除非指定交叉列表的行或列名称</span></p>
<p><span class="yiyi-st" id="yiyi-149">例如：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [65]: </span><span class="n">foo</span><span class="p">,</span> <span class="n">bar</span><span class="p">,</span> <span class="n">dull</span><span class="p">,</span> <span class="n">shiny</span><span class="p">,</span> <span class="n">one</span><span class="p">,</span> <span class="n">two</span> <span class="o">=</span> <span class="s1">'foo'</span><span class="p">,</span> <span class="s1">'bar'</span><span class="p">,</span> <span class="s1">'dull'</span><span class="p">,</span> <span class="s1">'shiny'</span><span class="p">,</span> <span class="s1">'one'</span><span class="p">,</span> <span class="s1">'two'</span>
<span class="gp">In [66]: </span><span class="n">a</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="n">foo</span><span class="p">,</span> <span class="n">foo</span><span class="p">,</span> <span class="n">bar</span><span class="p">,</span> <span class="n">bar</span><span class="p">,</span> <span class="n">foo</span><span class="p">,</span> <span class="n">foo</span><span class="p">],</span> <span class="n">dtype</span><span class="o">=</span><span class="nb">object</span><span class="p">)</span>
<span class="gp">In [67]: </span><span class="n">b</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="n">one</span><span class="p">,</span> <span class="n">one</span><span class="p">,</span> <span class="n">two</span><span class="p">,</span> <span class="n">one</span><span class="p">,</span> <span class="n">two</span><span class="p">,</span> <span class="n">one</span><span class="p">],</span> <span class="n">dtype</span><span class="o">=</span><span class="nb">object</span><span class="p">)</span>
<span class="gp">In [68]: </span><span class="n">c</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="n">dull</span><span class="p">,</span> <span class="n">dull</span><span class="p">,</span> <span class="n">shiny</span><span class="p">,</span> <span class="n">dull</span><span class="p">,</span> <span class="n">dull</span><span class="p">,</span> <span class="n">shiny</span><span class="p">],</span> <span class="n">dtype</span><span class="o">=</span><span class="nb">object</span><span class="p">)</span>
<span class="gp">In [69]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="p">[</span><span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">],</span> <span class="n">rownames</span><span class="o">=</span><span class="p">[</span><span class="s1">'a'</span><span class="p">],</span> <span class="n">colnames</span><span class="o">=</span><span class="p">[</span><span class="s1">'b'</span><span class="p">,</span> <span class="s1">'c'</span><span class="p">])</span>
<span class="gr">Out[69]: </span>
<span class="go">b one two </span>
<span class="go">c dull shiny dull shiny</span>
<span class="go">a </span>
<span class="go">bar 1 0 0 1</span>
<span class="go">foo 2 1 1 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-150">如果<code class="docutils literal"><span class="pre">crosstab</span></code>只接收两个Series，它将提供一个频率表。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [70]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'A'</span><span class="p">:</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">],</span> <span class="s1">'B'</span><span class="p">:</span> <span class="p">[</span><span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'C'</span><span class="p">:</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="n">np</span><span class="o">.</span><span class="n">nan</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [71]: </span><span class="n">df</span>
<span class="gr">Out[71]: </span>
<span class="go"> A B C</span>
<span class="go">0 1 3 1.0</span>
<span class="go">1 2 3 1.0</span>
<span class="go">2 2 4 NaN</span>
<span class="go">3 2 4 1.0</span>
<span class="go">4 2 4 1.0</span>
<span class="gp">In [72]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">df</span><span class="o">.</span><span class="n">A</span><span class="p">,</span> <span class="n">df</span><span class="o">.</span><span class="n">B</span><span class="p">)</span>
<span class="gr">Out[72]: </span>
<span class="go">B 3 4</span>
<span class="go">A </span>
<span class="go">1 1 0</span>
<span class="go">2 1 3</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-151">包含<code class="docutils literal"><span class="pre">Categorical</span></code>数据的任何输入都将包含在交叉列表中的类别中包含<strong>所有</strong>，即使实际数据不包含特定类别的任何实例。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [73]: </span><span class="n">foo</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">Categorical</span><span class="p">([</span><span class="s1">'a'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">],</span> <span class="n">categories</span><span class="o">=</span><span class="p">[</span><span class="s1">'a'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">,</span> <span class="s1">'c'</span><span class="p">])</span>
<span class="gp">In [74]: </span><span class="n">bar</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">Categorical</span><span class="p">([</span><span class="s1">'d'</span><span class="p">,</span> <span class="s1">'e'</span><span class="p">],</span> <span class="n">categories</span><span class="o">=</span><span class="p">[</span><span class="s1">'d'</span><span class="p">,</span> <span class="s1">'e'</span><span class="p">,</span> <span class="s1">'f'</span><span class="p">])</span>
<span class="gp">In [75]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">foo</span><span class="p">,</span> <span class="n">bar</span><span class="p">)</span>
<span class="gr">Out[75]: </span>
<span class="go">col_0 d e f</span>
<span class="go">row_0 </span>
<span class="go">a 1 0 0</span>
<span class="go">b 0 1 0</span>
<span class="go">c 0 0 0</span>
</pre></div>
</div>
<div class="section" id="normalization">
<h3><span class="yiyi-st" id="yiyi-152">Normalization</span></h3>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-153"><span class="versionmodified">版本0.18.1中的新功能。</span></span></p>
</div>
<p><span class="yiyi-st" id="yiyi-154">频率表也可以使用<code class="docutils literal"><span class="pre">normalize</span></code>参数进行标准化，以显示百分比而不是计数：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [76]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">df</span><span class="o">.</span><span class="n">A</span><span class="p">,</span> <span class="n">df</span><span class="o">.</span><span class="n">B</span><span class="p">,</span> <span class="n">normalize</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="gr">Out[76]: </span>
<span class="go">B 3 4</span>
<span class="go">A </span>
<span class="go">1 0.2 0.0</span>
<span class="go">2 0.2 0.6</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-155"><code class="docutils literal"><span class="pre">normalize</span></code>还可以标准化每行或每列中的值：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [77]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">df</span><span class="o">.</span><span class="n">A</span><span class="p">,</span> <span class="n">df</span><span class="o">.</span><span class="n">B</span><span class="p">,</span> <span class="n">normalize</span><span class="o">=</span><span class="s1">'columns'</span><span class="p">)</span>
<span class="gr">Out[77]: </span>
<span class="go">B 3 4</span>
<span class="go">A </span>
<span class="go">1 0.5 0.0</span>
<span class="go">2 0.5 1.0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-156"><code class="docutils literal"><span class="pre">crosstab</span></code>还可以传递第三个系列和聚合函数（<code class="docutils literal"><span class="pre">aggfunc</span></code>），将应用于由前两个系列定义的每个组中的第三个系列的值：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [78]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">df</span><span class="o">.</span><span class="n">A</span><span class="p">,</span> <span class="n">df</span><span class="o">.</span><span class="n">B</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="n">df</span><span class="o">.</span><span class="n">C</span><span class="p">,</span> <span class="n">aggfunc</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sum</span><span class="p">)</span>
<span class="gr">Out[78]: </span>
<span class="go">B 3 4</span>
<span class="go">A </span>
<span class="go">1 1.0 NaN</span>
<span class="go">2 1.0 2.0</span>
</pre></div>
</div>
</div>
<div class="section" id="id1">
<h3><span class="yiyi-st" id="yiyi-157">Adding Margins</span></h3>
<p><span class="yiyi-st" id="yiyi-158">最后，还可以添加边距或规范化此输出。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [79]: </span><span class="n">pd</span><span class="o">.</span><span class="n">crosstab</span><span class="p">(</span><span class="n">df</span><span class="o">.</span><span class="n">A</span><span class="p">,</span> <span class="n">df</span><span class="o">.</span><span class="n">B</span><span class="p">,</span> <span class="n">values</span><span class="o">=</span><span class="n">df</span><span class="o">.</span><span class="n">C</span><span class="p">,</span> <span class="n">aggfunc</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sum</span><span class="p">,</span> <span class="n">normalize</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">margins</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="gp"> ....:</span>
<span class="gr">Out[79]: </span>
<span class="go">B 3 4 All</span>
<span class="go">A </span>
<span class="go">1 0.25 0.0 0.25</span>
<span class="go">2 0.25 0.5 0.75</span>
<span class="go">All 0.50 0.5 1.00</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="tiling">
<span id="reshaping-tile-cut"></span><span id="reshaping-tile"></span><h2><span class="yiyi-st" id="yiyi-159">Tiling</span></h2>
<p><span class="yiyi-st" id="yiyi-160"><code class="docutils literal"><span class="pre">cut</span></code>函数计算输入数组值的分组，通常用于将连续变量转换为离散变量或分类变量：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [80]: </span><span class="n">ages</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">([</span><span class="mi">10</span><span class="p">,</span> <span class="mi">15</span><span class="p">,</span> <span class="mi">13</span><span class="p">,</span> <span class="mi">12</span><span class="p">,</span> <span class="mi">23</span><span class="p">,</span> <span class="mi">25</span><span class="p">,</span> <span class="mi">28</span><span class="p">,</span> <span class="mi">59</span><span class="p">,</span> <span class="mi">60</span><span class="p">])</span>
<span class="gp">In [81]: </span><span class="n">pd</span><span class="o">.</span><span class="n">cut</span><span class="p">(</span><span class="n">ages</span><span class="p">,</span> <span class="n">bins</span><span class="o">=</span><span class="mi">3</span><span class="p">)</span>
<span class="gr">Out[81]: </span>
<span class="go">[(9.95, 26.667], (9.95, 26.667], (9.95, 26.667], (9.95, 26.667], (9.95, 26.667], (9.95, 26.667], (26.667, 43.333], (43.333, 60], (43.333, 60]]</span>
<span class="go">Categories (3, object): [(9.95, 26.667] < (26.667, 43.333] < (43.333, 60]]</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-161">如果<code class="docutils literal"><span class="pre">bins</span></code>关键字是一个整数，那么将形成等宽字节。</span><span class="yiyi-st" id="yiyi-162">或者，我们可以指定自定义bin边：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [82]: </span><span class="n">pd</span><span class="o">.</span><span class="n">cut</span><span class="p">(</span><span class="n">ages</span><span class="p">,</span> <span class="n">bins</span><span class="o">=</span><span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">18</span><span class="p">,</span> <span class="mi">35</span><span class="p">,</span> <span class="mi">70</span><span class="p">])</span>
<span class="gr">Out[82]: </span>
<span class="go">[(0, 18], (0, 18], (0, 18], (0, 18], (18, 35], (18, 35], (18, 35], (35, 70], (35, 70]]</span>
<span class="go">Categories (3, object): [(0, 18] < (18, 35] < (35, 70]]</span>
</pre></div>
</div>
</div>
<div class="section" id="computing-indicator-dummy-variables">
<span id="reshaping-dummies"></span><h2><span class="yiyi-st" id="yiyi-163">Computing indicator / dummy variables</span></h2>
<p><span class="yiyi-st" id="yiyi-164">为了将分类变量转换为“虚拟”或“指示符”DataFrame，例如具有<code class="docutils literal"><span class="pre">k</span></code>不同值的DataFrame（a Series）中的列，可以导出包含<code class="docutils literal"><span class="pre">k</span></code></span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [83]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'key'</span><span class="p">:</span> <span class="nb">list</span><span class="p">(</span><span class="s1">'bbacab'</span><span class="p">),</span> <span class="s1">'data1'</span><span class="p">:</span> <span class="nb">range</span><span class="p">(</span><span class="mi">6</span><span class="p">)})</span>
<span class="gp">In [84]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">[</span><span class="s1">'key'</span><span class="p">])</span>
<span class="gr">Out[84]: </span>
<span class="go"> a b c</span>
<span class="go">0 0 1 0</span>
<span class="go">1 0 1 0</span>
<span class="go">2 1 0 0</span>
<span class="go">3 0 0 1</span>
<span class="go">4 1 0 0</span>
<span class="go">5 0 1 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-165">有时，使用列名称前缀是有用的，例如在将结果与原始DataFrame合并时：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [85]: </span><span class="n">dummies</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">[</span><span class="s1">'key'</span><span class="p">],</span> <span class="n">prefix</span><span class="o">=</span><span class="s1">'key'</span><span class="p">)</span>
<span class="gp">In [86]: </span><span class="n">dummies</span>
<span class="gr">Out[86]: </span>
<span class="go"> key_a key_b key_c</span>
<span class="go">0 0 1 0</span>
<span class="go">1 0 1 0</span>
<span class="go">2 1 0 0</span>
<span class="go">3 0 0 1</span>
<span class="go">4 1 0 0</span>
<span class="go">5 0 1 0</span>
<span class="gp">In [87]: </span><span class="n">df</span><span class="p">[[</span><span class="s1">'data1'</span><span class="p">]]</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">dummies</span><span class="p">)</span>
<span class="gr">Out[87]: </span>
<span class="go"> data1 key_a key_b key_c</span>
<span class="go">0 0 0 1 0</span>
<span class="go">1 1 0 1 0</span>
<span class="go">2 2 1 0 0</span>
<span class="go">3 3 0 0 1</span>
<span class="go">4 4 1 0 0</span>
<span class="go">5 5 0 1 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-166">此函数通常与<code class="docutils literal"><span class="pre">cut</span></code>等离散函数一起使用：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [88]: </span><span class="n">values</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
<span class="gp">In [89]: </span><span class="n">values</span>
<span class="gr">Out[89]: </span>
<span class="go">array([ 0.4082, -1.0481, -0.0257, -0.9884, 0.0941, 1.2627, 1.29 ,</span>
<span class="go"> 0.0824, -0.0558, 0.5366])</span>
<span class="gp">In [90]: </span><span class="n">bins</span> <span class="o">=</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mf">0.2</span><span class="p">,</span> <span class="mf">0.4</span><span class="p">,</span> <span class="mf">0.6</span><span class="p">,</span> <span class="mf">0.8</span><span class="p">,</span> <span class="mi">1</span><span class="p">]</span>
<span class="gp">In [91]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">pd</span><span class="o">.</span><span class="n">cut</span><span class="p">(</span><span class="n">values</span><span class="p">,</span> <span class="n">bins</span><span class="p">))</span>
<span class="gr">Out[91]: </span>
<span class="go"> (0, 0.2] (0.2, 0.4] (0.4, 0.6] (0.6, 0.8] (0.8, 1]</span>
<span class="go">0 0 0 1 0 0</span>
<span class="go">1 0 0 0 0 0</span>
<span class="go">2 0 0 0 0 0</span>
<span class="go">3 0 0 0 0 0</span>
<span class="go">4 1 0 0 0 0</span>
<span class="go">5 0 0 0 0 0</span>
<span class="go">6 0 0 0 0 0</span>
<span class="go">7 1 0 0 0 0</span>
<span class="go">8 0 0 0 0 0</span>
<span class="go">9 0 0 1 0 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-167">另请参见<a class="reference internal" href="generated/pandas.Series.str.get_dummies.html#pandas.Series.str.get_dummies" title="pandas.Series.str.get_dummies"><code class="xref py py-func docutils literal"><span class="pre">Series.str.get_dummies</span></code></a>。</span></p>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-168"><span class="versionmodified">版本0.15.0中的新功能。</span></span></p>
</div>
<p><span class="yiyi-st" id="yiyi-169"><a class="reference internal" href="generated/pandas.get_dummies.html#pandas.get_dummies" title="pandas.get_dummies"><code class="xref py py-func docutils literal"><span class="pre">get_dummies()</span></code></a>也接受一个DataFrame。</span><span class="yiyi-st" id="yiyi-170">默认情况下，所有类别变量（在统计学意义上为分类变量，具有<cite>对象</cite>或<cite>分类 t> dtype）被编码为虚拟变量。</cite></span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [92]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'A'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'a'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">,</span> <span class="s1">'a'</span><span class="p">],</span> <span class="s1">'B'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'c'</span><span class="p">,</span> <span class="s1">'c'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'C'</span><span class="p">:</span> <span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [93]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">)</span>
<span class="gr">Out[93]: </span>
<span class="go"> C A_a A_b B_b B_c</span>
<span class="go">0 1 1 0 0 1</span>
<span class="go">1 2 0 1 0 1</span>
<span class="go">2 3 1 0 1 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-171">在输出中包含所有非对象列。</span></p>
<p><span class="yiyi-st" id="yiyi-172">您可以控制使用<code class="docutils literal"><span class="pre">columns</span></code>关键字编码的列。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [94]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'A'</span><span class="p">])</span>
<span class="gr">Out[94]: </span>
<span class="go"> B C A_a A_b</span>
<span class="go">0 c 1 1 0</span>
<span class="go">1 c 2 0 1</span>
<span class="go">2 b 3 1 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-173">请注意，<code class="docutils literal"><span class="pre">B</span></code>列仍包含在输出中，它只是没有被编码。</span><span class="yiyi-st" id="yiyi-174">如果您不想将其包含在输出中，则可以在调用<code class="docutils literal"><span class="pre">get_dummies</span></code>之前拖动<code class="docutils literal"><span class="pre">B</span></code>。</span></p>
<p><span class="yiyi-st" id="yiyi-175">与Series版本一样，您可以传递<code class="docutils literal"><span class="pre">prefix</span></code>和<code class="docutils literal"><span class="pre">prefix_sep</span></code>的值。</span><span class="yiyi-st" id="yiyi-176">默认情况下，列名称用作前缀，“_”用作前缀分隔符。</span><span class="yiyi-st" id="yiyi-177">您可以通过3种方式指定<code class="docutils literal"><span class="pre">prefix</span></code>和<code class="docutils literal"><span class="pre">prefix_sep</span></code></span></p>
<ul class="simple">
<li><span class="yiyi-st" id="yiyi-178">string：对要编码的每个列，使用<code class="docutils literal"><span class="pre">prefix</span></code>或<code class="docutils literal"><span class="pre">prefix_sep</span></code>的相同值</span></li>
<li><span class="yiyi-st" id="yiyi-179">list：必须与正在编码的列的长度相同。</span></li>
<li><span class="yiyi-st" id="yiyi-180">dict：将列名映射到前缀</span></li>
</ul>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [95]: </span><span class="n">simple</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">prefix</span><span class="o">=</span><span class="s1">'new_prefix'</span><span class="p">)</span>
<span class="gp">In [96]: </span><span class="n">simple</span>
<span class="gr">Out[96]: </span>
<span class="go"> C new_prefix_a new_prefix_b new_prefix_b new_prefix_c</span>
<span class="go">0 1 1 0 0 1</span>
<span class="go">1 2 0 1 0 1</span>
<span class="go">2 3 1 0 1 0</span>
<span class="gp">In [97]: </span><span class="n">from_list</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">prefix</span><span class="o">=</span><span class="p">[</span><span class="s1">'from_A'</span><span class="p">,</span> <span class="s1">'from_B'</span><span class="p">])</span>
<span class="gp">In [98]: </span><span class="n">from_list</span>
<span class="gr">Out[98]: </span>
<span class="go"> C from_A_a from_A_b from_B_b from_B_c</span>
<span class="go">0 1 1 0 0 1</span>
<span class="go">1 2 0 1 0 1</span>
<span class="go">2 3 1 0 1 0</span>
<span class="gp">In [99]: </span><span class="n">from_dict</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">prefix</span><span class="o">=</span><span class="p">{</span><span class="s1">'B'</span><span class="p">:</span> <span class="s1">'from_B'</span><span class="p">,</span> <span class="s1">'A'</span><span class="p">:</span> <span class="s1">'from_A'</span><span class="p">})</span>
<span class="gp">In [100]: </span><span class="n">from_dict</span>
<span class="gr">Out[100]: </span>
<span class="go"> C from_A_a from_A_b from_B_b from_B_c</span>
<span class="go">0 1 1 0 0 1</span>
<span class="go">1 2 0 1 0 1</span>
<span class="go">2 3 1 0 1 0</span>
</pre></div>
</div>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-181"><span class="versionmodified">版本0.18.0中的新功能。</span></span></p>
</div>
<p><span class="yiyi-st" id="yiyi-182">有时，在将结果馈送到统计模型时，仅保留k-1级别的分类变量以避免共线性将是有用的。</span><span class="yiyi-st" id="yiyi-183">您可以通过打开<code class="docutils literal"><span class="pre">drop_first</span></code>切换到此模式。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [101]: </span><span class="n">s</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">Series</span><span class="p">(</span><span class="nb">list</span><span class="p">(</span><span class="s1">'abcaa'</span><span class="p">))</span>
<span class="gp">In [102]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">s</span><span class="p">)</span>
<span class="gr">Out[102]: </span>
<span class="go"> a b c</span>
<span class="go">0 1 0 0</span>
<span class="go">1 0 1 0</span>
<span class="go">2 0 0 1</span>
<span class="go">3 1 0 0</span>
<span class="go">4 1 0 0</span>
<span class="gp">In [103]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">s</span><span class="p">,</span> <span class="n">drop_first</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="gr">Out[103]: </span>
<span class="go"> b c</span>
<span class="go">0 0 0</span>
<span class="go">1 1 0</span>
<span class="go">2 0 1</span>
<span class="go">3 0 0</span>
<span class="go">4 0 0</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-184">当列只包含一个级别时，将在结果中省略。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [104]: </span><span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">({</span><span class="s1">'A'</span><span class="p">:</span><span class="nb">list</span><span class="p">(</span><span class="s1">'aaaaa'</span><span class="p">),</span><span class="s1">'B'</span><span class="p">:</span><span class="nb">list</span><span class="p">(</span><span class="s1">'ababc'</span><span class="p">)})</span>
<span class="gp">In [105]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">)</span>
<span class="gr">Out[105]: </span>
<span class="go"> A_a B_a B_b B_c</span>
<span class="go">0 1 1 0 0</span>
<span class="go">1 1 0 1 0</span>
<span class="go">2 1 1 0 0</span>
<span class="go">3 1 0 1 0</span>
<span class="go">4 1 0 0 1</span>
<span class="gp">In [106]: </span><span class="n">pd</span><span class="o">.</span><span class="n">get_dummies</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">drop_first</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="gr">Out[106]: </span>
<span class="go"> B_b B_c</span>
<span class="go">0 0 0</span>