forked from wizardforcel/pandas-doc-zh
-
Notifications
You must be signed in to change notification settings - Fork 2
/
merging.html
923 lines (881 loc) · 138 KB
/
merging.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
<span id="merging"></span><h1><span class="yiyi-st" id="yiyi-67">Merge, join, and concatenate</span></h1>
<blockquote>
<p>原文：<a href="http://pandas.pydata.org/pandas-docs/stable/merging.html">http://pandas.pydata.org/pandas-docs/stable/merging.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>
<p><span class="yiyi-st" id="yiyi-68">pandas提供了各种设施，以便在连接/合并类型操作的情况下，轻松地将Series，DataFrame和Panel对象与索引的各种集合逻辑以及关系代数功能组合在一起。</span></p>
<div class="section" id="concatenating-objects">
<span id="merging-concat"></span><h2><span class="yiyi-st" id="yiyi-69">Concatenating objects</span></h2>
<p><span class="yiyi-st" id="yiyi-70"><code class="docutils literal"><span class="pre">concat</span></code>函数（在主pandas命名空间中）执行沿轴执行连接操作的所有繁重工作，同时执行索引（如果有）的可选集逻辑（联合或交集）轴。</span><span class="yiyi-st" id="yiyi-71">注意，我说“如果有”，因为对于Series只有一个可能的级联轴。</span></p>
<p><span class="yiyi-st" id="yiyi-72">在介绍<code class="docutils literal"><span class="pre">concat</span></code>的所有细节以及它能做什么之前，这里有一个简单的例子：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [1]: </span><span class="n">df1</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">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</span><span class="p">,</span> <span class="s1">'A3'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">,</span> <span class="s1">'B3'</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">'C0'</span><span class="p">,</span> <span class="s1">'C1'</span><span class="p">,</span> <span class="s1">'C2'</span><span class="p">,</span> <span class="s1">'C3'</span><span class="p">],</span>
<span class="gp"> ...:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D1'</span><span class="p">,</span> <span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">]},</span>
<span class="gp"> ...:</span> <span class="n">index</span><span class="o">=</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">3</span><span class="p">])</span>
<span class="gp"> ...:</span>
<span class="gp">In [2]: </span><span class="n">df2</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">'A4'</span><span class="p">,</span> <span class="s1">'A5'</span><span class="p">,</span> <span class="s1">'A6'</span><span class="p">,</span> <span class="s1">'A7'</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">'B4'</span><span class="p">,</span> <span class="s1">'B5'</span><span class="p">,</span> <span class="s1">'B6'</span><span class="p">,</span> <span class="s1">'B7'</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">'C4'</span><span class="p">,</span> <span class="s1">'C5'</span><span class="p">,</span> <span class="s1">'C6'</span><span class="p">,</span> <span class="s1">'C7'</span><span class="p">],</span>
<span class="gp"> ...:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D4'</span><span class="p">,</span> <span class="s1">'D5'</span><span class="p">,</span> <span class="s1">'D6'</span><span class="p">,</span> <span class="s1">'D7'</span><span class="p">]},</span>
<span class="gp"> ...:</span> <span class="n">index</span><span class="o">=</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">6</span><span class="p">,</span> <span class="mi">7</span><span class="p">])</span>
<span class="gp"> ...:</span>
<span class="gp">In [3]: </span><span class="n">df3</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">'A8'</span><span class="p">,</span> <span class="s1">'A9'</span><span class="p">,</span> <span class="s1">'A10'</span><span class="p">,</span> <span class="s1">'A11'</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">'B8'</span><span class="p">,</span> <span class="s1">'B9'</span><span class="p">,</span> <span class="s1">'B10'</span><span class="p">,</span> <span class="s1">'B11'</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">'C8'</span><span class="p">,</span> <span class="s1">'C9'</span><span class="p">,</span> <span class="s1">'C10'</span><span class="p">,</span> <span class="s1">'C11'</span><span class="p">],</span>
<span class="gp"> ...:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D8'</span><span class="p">,</span> <span class="s1">'D9'</span><span class="p">,</span> <span class="s1">'D10'</span><span class="p">,</span> <span class="s1">'D11'</span><span class="p">]},</span>
<span class="gp"> ...:</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="mi">8</span><span class="p">,</span> <span class="mi">9</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="mi">11</span><span class="p">])</span>
<span class="gp"> ...:</span>
<span class="gp">In [4]: </span><span class="n">frames</span> <span class="o">=</span> <span class="p">[</span><span class="n">df1</span><span class="p">,</span> <span class="n">df2</span><span class="p">,</span> <span class="n">df3</span><span class="p">]</span>
<span class="gp">In [5]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">frames</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_basic.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_basic.png">
<p><span class="yiyi-st" id="yiyi-73">像它在ndarrays上的同级函数一样，<code class="docutils literal"><span class="pre">numpy.concatenate</span></code>，<code class="docutils literal"><span class="pre">pandas.concat</span></code>接受同类型对象的列表或dict，并将它们与“与其他轴“：</span></p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">objs</span><span class="p">,</span> <span class="n">axis</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">join</span><span class="o">=</span><span class="s1">'outer'</span><span class="p">,</span> <span class="n">join_axes</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">ignore_index</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span>
<span class="n">keys</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">levels</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">names</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">verify_integrity</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span>
<span class="n">copy</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
</pre></div>
</div>
<ul class="simple">
<li><span class="yiyi-st" id="yiyi-74"><code class="docutils literal"><span class="pre">objs</span></code>：Series，DataFrame或Panel对象的序列或映射。</span><span class="yiyi-st" id="yiyi-75">如果传递了dict，则排序的键将用作<cite>键</cite>参数，除非它被传递，在这种情况下，将选择值（见下文）。</span><span class="yiyi-st" id="yiyi-76">任何无对象将被静默删除，除非它们都是无，在这种情况下将引发一个ValueError。</span></li>
<li><span class="yiyi-st" id="yiyi-77"><code class="docutils literal"><span class="pre">axis</span></code>：{0,1，...}，默认为0。</span><span class="yiyi-st" id="yiyi-78">沿着连接的轴。</span></li>
<li><span class="yiyi-st" id="yiyi-79"><code class="docutils literal"><span class="pre">join</span></code>：{'inner'，'outer'}，默认为“outer”。</span><span class="yiyi-st" id="yiyi-80">如何处理其他轴上的索引。</span><span class="yiyi-st" id="yiyi-81">outer为联合和inner为交集。</span></li>
<li><span class="yiyi-st" id="yiyi-82"><code class="docutils literal"><span class="pre">ignore_index</span></code>：boolean，default False。</span><span class="yiyi-st" id="yiyi-83">如果为True，请不要使用并置轴上的索引值。</span><span class="yiyi-st" id="yiyi-84">结果轴将被标记为0，...，n-1。</span><span class="yiyi-st" id="yiyi-85">如果要连接其中并置轴没有有意义的索引信息的对象，这将非常有用。</span><span class="yiyi-st" id="yiyi-86">注意，其他轴上的索引值在连接中仍然受到尊重。</span></li>
<li><span class="yiyi-st" id="yiyi-87"><code class="docutils literal"><span class="pre">join_axes</span></code>：Index对象列表。</span><span class="yiyi-st" id="yiyi-88">用于其他n-1轴的特定索引，而不是执行内部/外部设置逻辑。</span></li>
<li><span class="yiyi-st" id="yiyi-89"><code class="docutils literal"><span class="pre">keys</span></code>：序列，默认值无。</span><span class="yiyi-st" id="yiyi-90">使用传递的键作为最外层构建层次索引。</span><span class="yiyi-st" id="yiyi-91">如果为多索引，应该使用元组。</span></li>
<li><span class="yiyi-st" id="yiyi-92"><code class="docutils literal"><span class="pre">levels</span></code>：序列列表，默认值无。</span><span class="yiyi-st" id="yiyi-93">用于构建MultiIndex的特定级别（唯一值）。</span><span class="yiyi-st" id="yiyi-94">否则，它们将从键推断。</span></li>
<li><span class="yiyi-st" id="yiyi-95"><code class="docutils literal"><span class="pre">names</span></code>：list，default无。</span><span class="yiyi-st" id="yiyi-96">结果层次索引中的级别的名称。</span></li>
<li><span class="yiyi-st" id="yiyi-97"><code class="docutils literal"><span class="pre">verify_integrity</span></code>：boolean，default False。</span><span class="yiyi-st" id="yiyi-98">检查新连接的轴是否包含重复项。</span><span class="yiyi-st" id="yiyi-99">这相对于实际的数据串联可能是非常昂贵的。</span></li>
<li><span class="yiyi-st" id="yiyi-100"><code class="docutils literal"><span class="pre">copy</span></code>：boolean，default True。</span><span class="yiyi-st" id="yiyi-101">如果为False，请勿不必要地复制数据。</span></li>
</ul>
<p><span class="yiyi-st" id="yiyi-102">没有一点点上下文和例子许多这些参数没有多大意义。</span><span class="yiyi-st" id="yiyi-103">让我们来看上面的例子。</span><span class="yiyi-st" id="yiyi-104">假设我们想要将特定的键与每一个被切碎的DataFrame关联起来。</span><span class="yiyi-st" id="yiyi-105">我们可以使用<code class="docutils literal"><span class="pre">keys</span></code>参数：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [6]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">frames</span><span class="p">,</span> <span class="n">keys</span><span class="o">=</span><span class="p">[</span><span class="s1">'x'</span><span class="p">,</span> <span class="s1">'y'</span><span class="p">,</span> <span class="s1">'z'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_keys.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_keys.png">
<p><span class="yiyi-st" id="yiyi-106">正如你可以看到的（如果你已经阅读了文档的其余部分），结果对象的索引具有<a class="reference internal" href="advanced.html#advanced-hierarchical"><span class="std std-ref">hierarchical index</span></a>。</span><span class="yiyi-st" id="yiyi-107">这意味着我们现在可以做的东西，像通过键选择每个块：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [7]: </span><span class="n">result</span><span class="o">.</span><span class="n">ix</span><span class="p">[</span><span class="s1">'y'</span><span class="p">]</span>
<span class="gr">Out[7]: </span>
<span class="go"> A B C D</span>
<span class="go">4 A4 B4 C4 D4</span>
<span class="go">5 A5 B5 C5 D5</span>
<span class="go">6 A6 B6 C6 D6</span>
<span class="go">7 A7 B7 C7 D7</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-108">这不是一个伸展，看看这可以非常有用。</span><span class="yiyi-st" id="yiyi-109">有关此功能的更多详细信息。</span></p>
<div class="admonition note">
<p class="first admonition-title"><span class="yiyi-st" id="yiyi-110">注意</span></p>
<p class="last"><span class="yiyi-st" id="yiyi-111">然而，值得注意的是，<code class="docutils literal"><span class="pre">concat</span></code>（因此<code class="docutils literal"><span class="pre">append</span></code>）会创建数据的完整副本，并且不断重用此函数可能会产生重大的性能损失。</span><span class="yiyi-st" id="yiyi-112">如果需要使用对多个数据集的操作，请使用列表推导。</span></p>
</div>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">frames</span> <span class="o">=</span> <span class="p">[</span> <span class="n">process_your_file</span><span class="p">(</span><span class="n">f</span><span class="p">)</span> <span class="k">for</span> <span class="n">f</span> <span class="ow">in</span> <span class="n">files</span> <span class="p">]</span>
<span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">frames</span><span class="p">)</span>
</pre></div>
</div>
<div class="section" id="set-logic-on-the-other-axes">
<h3><span class="yiyi-st" id="yiyi-113">Set logic on the other axes</span></h3>
<p><span class="yiyi-st" id="yiyi-114">例如，当将多个DataFrames（或面板或...）粘合在一起时，您可以选择如何处理其他轴（不是并置的轴）。</span><span class="yiyi-st" id="yiyi-115">这可以通过三种方式完成：</span></p>
<ul class="simple">
<li><span class="yiyi-st" id="yiyi-116">取它们的（排序）并集，<code class="docutils literal"><span class="pre">join='outer'</span></code>。</span><span class="yiyi-st" id="yiyi-117">这是默认选项，因为它导致零信息丢失。</span></li>
<li><span class="yiyi-st" id="yiyi-118">以交叉点<code class="docutils literal"><span class="pre">join='inner'</span></code>。</span></li>
<li><span class="yiyi-st" id="yiyi-119">使用特定索引（在DataFrame的情况下）或索引（在Panel或未来更高维度的对象的情况下），即<code class="docutils literal"><span class="pre">join_axes</span></code>参数</span></li>
</ul>
<p><span class="yiyi-st" id="yiyi-120">这里是每个这些方法的示例。</span><span class="yiyi-st" id="yiyi-121">首先，默认的<code class="docutils literal"><span class="pre">join='outer'</span></code>行为：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [8]: </span><span class="n">df4</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">'B'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'B2'</span><span class="p">,</span> <span class="s1">'B3'</span><span class="p">,</span> <span class="s1">'B6'</span><span class="p">,</span> <span class="s1">'B7'</span><span class="p">],</span>
<span class="gp"> ...:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">,</span> <span class="s1">'D6'</span><span class="p">,</span> <span class="s1">'D7'</span><span class="p">],</span>
<span class="gp"> ...:</span> <span class="s1">'F'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'F2'</span><span class="p">,</span> <span class="s1">'F3'</span><span class="p">,</span> <span class="s1">'F6'</span><span class="p">,</span> <span class="s1">'F7'</span><span class="p">]},</span>
<span class="gp"> ...:</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">7</span><span class="p">])</span>
<span class="gp"> ...:</span>
<span class="gp">In [9]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">df4</span><span class="p">],</span> <span class="n">axis</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_axis1.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_axis1.png">
<p><span class="yiyi-st" id="yiyi-122">注意，行索引已经被组合和排序。</span><span class="yiyi-st" id="yiyi-123">这与<code class="docutils literal"><span class="pre">join='inner'</span></code>是一样的：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [10]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">df4</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="n">join</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_axis1_inner.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_axis1_inner.png">
<p><span class="yiyi-st" id="yiyi-124">最后，假设我们只想从原始DataFrame重用<em>确切索引</em>：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [11]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">df4</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="n">join_axes</span><span class="o">=</span><span class="p">[</span><span class="n">df1</span><span class="o">.</span><span class="n">index</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_axis1_join_axes.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_axis1_join_axes.png">
</div>
<div class="section" id="concatenating-using-append">
<span id="merging-concatenation"></span><h3><span class="yiyi-st" id="yiyi-125">Concatenating using <code class="docutils literal"><span class="pre">append</span></code></span></h3>
<p><span class="yiyi-st" id="yiyi-126">对<code class="docutils literal"><span class="pre">concat</span></code>有用的快捷方式是Series和DataFrame上的<code class="docutils literal"><span class="pre">append</span></code>实例方法。</span><span class="yiyi-st" id="yiyi-127">这些方法实际上早于<code class="docutils literal"><span class="pre">concat</span></code>。</span><span class="yiyi-st" id="yiyi-128">它们沿<code class="docutils literal"><span class="pre">axis=0</span></code>连接，即索引：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [12]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">df2</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append1.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append1.png">
<p><span class="yiyi-st" id="yiyi-129">在DataFrame的情况下，索引必须是不相交的，但列不需要是：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [13]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">df4</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append2.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append2.png">
<p><span class="yiyi-st" id="yiyi-130"><code class="docutils literal"><span class="pre">append</span></code>可能需要多个对象进行连接：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [14]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">append</span><span class="p">([</span><span class="n">df2</span><span class="p">,</span> <span class="n">df3</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append3.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append3.png">
<div class="admonition note">
<p class="first admonition-title"><span class="yiyi-st" id="yiyi-131">注意</span></p>
<p class="last"><span class="yiyi-st" id="yiyi-132">与不附加到原始列表并不返回任何内容的<cite>list.append</cite>方法不同，<code class="docutils literal"><span class="pre">append</span></code> <strong>不会</strong>修改<code class="docutils literal"><span class="pre">df1</span></code>并返回其附带<code class="docutils literal"><span class="pre">df2</span></code>的副本。</span></p>
</div>
</div>
<div class="section" id="ignoring-indexes-on-the-concatenation-axis">
<span id="merging-ignore-index"></span><h3><span class="yiyi-st" id="yiyi-133">Ignoring indexes on the concatenation axis</span></h3>
<p><span class="yiyi-st" id="yiyi-134">对于没有有意义索引的DataFrames，您可能希望附加它们，并忽略它们可能具有重叠索引的事实：</span></p>
<p><span class="yiyi-st" id="yiyi-135">为此，请使用<code class="docutils literal"><span class="pre">ignore_index</span></code>参数：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [15]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">df4</span><span class="p">],</span> <span class="n">ignore_index</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_ignore_index.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_ignore_index.png">
<p><span class="yiyi-st" id="yiyi-136">这也是<code class="docutils literal"><span class="pre">DataFrame.append</span></code>的有效参数：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [16]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">df4</span><span class="p">,</span> <span class="n">ignore_index</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append_ignore_index.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append_ignore_index.png">
</div>
<div class="section" id="concatenating-with-mixed-ndims">
<span id="merging-mixed-ndims"></span><h3><span class="yiyi-st" id="yiyi-137">Concatenating with mixed ndims</span></h3>
<p><span class="yiyi-st" id="yiyi-138">您可以连接Series和DataFrames的混合。</span><span class="yiyi-st" id="yiyi-139">该系列将被转换为DataFrames，列名称为Series的名称。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [17]: </span><span class="n">s1</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="s1">'X0'</span><span class="p">,</span> <span class="s1">'X1'</span><span class="p">,</span> <span class="s1">'X2'</span><span class="p">,</span> <span class="s1">'X3'</span><span class="p">],</span> <span class="n">name</span><span class="o">=</span><span class="s1">'X'</span><span class="p">)</span>
<span class="gp">In [18]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">s1</span><span class="p">],</span> <span class="n">axis</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_mixed_ndim.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_mixed_ndim.png">
<p><span class="yiyi-st" id="yiyi-140">如果未命名的系列通过，它们将被连续编号。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [19]: </span><span class="n">s2</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="s1">'_0'</span><span class="p">,</span> <span class="s1">'_1'</span><span class="p">,</span> <span class="s1">'_2'</span><span class="p">,</span> <span class="s1">'_3'</span><span class="p">])</span>
<span class="gp">In [20]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">s2</span><span class="p">,</span> <span class="n">s2</span><span class="p">,</span> <span class="n">s2</span><span class="p">],</span> <span class="n">axis</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_unnamed_series.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_unnamed_series.png">
<p><span class="yiyi-st" id="yiyi-141">传递<code class="docutils literal"><span class="pre">ignore_index=True</span></code>将删除所有名称引用。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [21]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">df1</span><span class="p">,</span> <span class="n">s1</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="n">ignore_index</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_series_ignore_index.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_series_ignore_index.png">
</div>
<div class="section" id="more-concatenating-with-group-keys">
<h3><span class="yiyi-st" id="yiyi-142">More concatenating with group keys</span></h3>
<p><span class="yiyi-st" id="yiyi-143"><code class="docutils literal"><span class="pre">keys</span></code>参数的常见用法是在基于现有系列创建新的DataFrame时覆盖列名。</span><span class="yiyi-st" id="yiyi-144">请注意默认行为是如何让结果DataFrame继承父系列名称（如果存在）。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [22]: </span><span class="n">s3</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="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">3</span><span class="p">],</span> <span class="n">name</span><span class="o">=</span><span class="s1">'foo'</span><span class="p">)</span>
<span class="gp">In [23]: </span><span class="n">s4</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="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">3</span><span class="p">])</span>
<span class="gp">In [24]: </span><span class="n">s5</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="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">5</span><span class="p">])</span>
<span class="gp">In [25]: </span><span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">s3</span><span class="p">,</span> <span class="n">s4</span><span class="p">,</span> <span class="n">s5</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="gr">Out[25]: </span>
<span class="go"> foo 0 1</span>
<span class="go">0 0 0 0</span>
<span class="go">1 1 1 1</span>
<span class="go">2 2 2 4</span>
<span class="go">3 3 3 5</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-145">通过<code class="docutils literal"><span class="pre">keys</span></code>参数，我们可以覆盖现有的列名。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [26]: </span><span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">([</span><span class="n">s3</span><span class="p">,</span> <span class="n">s4</span><span class="p">,</span> <span class="n">s5</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="n">keys</span><span class="o">=</span><span class="p">[</span><span class="s1">'red'</span><span class="p">,</span><span class="s1">'blue'</span><span class="p">,</span><span class="s1">'yellow'</span><span class="p">])</span>
<span class="gr">Out[26]: </span>
<span class="go"> red blue yellow</span>
<span class="go">0 0 0 0</span>
<span class="go">1 1 1 1</span>
<span class="go">2 2 2 4</span>
<span class="go">3 3 3 5</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-146">让我们现在考虑一个变化的第一个例子：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [27]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">frames</span><span class="p">,</span> <span class="n">keys</span><span class="o">=</span><span class="p">[</span><span class="s1">'x'</span><span class="p">,</span> <span class="s1">'y'</span><span class="p">,</span> <span class="s1">'z'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_group_keys2.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_group_keys2.png">
<p><span class="yiyi-st" id="yiyi-147">您还可以将dict传递到<code class="docutils literal"><span class="pre">concat</span></code>，在这种情况下，dict键将用于<code class="docutils literal"><span class="pre">keys</span></code>参数（除非指定了其他键）：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [28]: </span><span class="n">pieces</span> <span class="o">=</span> <span class="p">{</span><span class="s1">'x'</span><span class="p">:</span> <span class="n">df1</span><span class="p">,</span> <span class="s1">'y'</span><span class="p">:</span> <span class="n">df2</span><span class="p">,</span> <span class="s1">'z'</span><span class="p">:</span> <span class="n">df3</span><span class="p">}</span>
<span class="gp">In [29]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">pieces</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_dict.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_dict.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [30]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">pieces</span><span class="p">,</span> <span class="n">keys</span><span class="o">=</span><span class="p">[</span><span class="s1">'z'</span><span class="p">,</span> <span class="s1">'y'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_dict_keys.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_dict_keys.png">
<p><span class="yiyi-st" id="yiyi-148">创建的MultiIndex具有根据传递的键和DataFrame段的索引构建的级别：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [31]: </span><span class="n">result</span><span class="o">.</span><span class="n">index</span><span class="o">.</span><span class="n">levels</span>
<span class="gr">Out[31]: </span><span class="n">FrozenList</span><span class="p">([[</span><span class="s1">u'z'</span><span class="p">,</span> <span class="s1">u'y'</span><span class="p">],</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">6</span><span class="p">,</span> <span class="mi">7</span><span class="p">,</span> <span class="mi">8</span><span class="p">,</span> <span class="mi">9</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="mi">11</span><span class="p">]])</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-149">如果您想指定其他级别（偶尔会这样），您可以使用<code class="docutils literal"><span class="pre">levels</span></code>参数：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [32]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">concat</span><span class="p">(</span><span class="n">pieces</span><span class="p">,</span> <span class="n">keys</span><span class="o">=</span><span class="p">[</span><span class="s1">'x'</span><span class="p">,</span> <span class="s1">'y'</span><span class="p">,</span> <span class="s1">'z'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="n">levels</span><span class="o">=</span><span class="p">[[</span><span class="s1">'z'</span><span class="p">,</span> <span class="s1">'y'</span><span class="p">,</span> <span class="s1">'x'</span><span class="p">,</span> <span class="s1">'w'</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">'group_key'</span><span class="p">])</span>
<span class="gp"> ....:</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_dict_keys_names.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_concat_dict_keys_names.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [33]: </span><span class="n">result</span><span class="o">.</span><span class="n">index</span><span class="o">.</span><span class="n">levels</span>
<span class="gr">Out[33]: </span><span class="n">FrozenList</span><span class="p">([[</span><span class="s1">u'z'</span><span class="p">,</span> <span class="s1">u'y'</span><span class="p">,</span> <span class="s1">u'x'</span><span class="p">,</span> <span class="s1">u'w'</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">2</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">5</span><span class="p">,</span> <span class="mi">6</span><span class="p">,</span> <span class="mi">7</span><span class="p">,</span> <span class="mi">8</span><span class="p">,</span> <span class="mi">9</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="mi">11</span><span class="p">]])</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-150">是的，这是相当深奥，但实际上是实现像GroupBy，其中分类变量的顺序是有意义的。</span></p>
</div>
<div class="section" id="appending-rows-to-a-dataframe">
<span id="merging-append-row"></span><h3><span class="yiyi-st" id="yiyi-151">Appending rows to a DataFrame</span></h3>
<p><span class="yiyi-st" id="yiyi-152">虽然不是特别有效（因为必须创建一个新的对象），你可以通过传递一个Series或dict到<code class="docutils literal"><span class="pre">append</span></code>，它返回一个新的DataFrame如上所示，附加一行到DataFrame。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [34]: </span><span class="n">s2</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="s1">'X0'</span><span class="p">,</span> <span class="s1">'X1'</span><span class="p">,</span> <span class="s1">'X2'</span><span class="p">,</span> <span class="s1">'X3'</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="s1">'C'</span><span class="p">,</span> <span class="s1">'D'</span><span class="p">])</span>
<span class="gp">In [35]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">s2</span><span class="p">,</span> <span class="n">ignore_index</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append_series_as_row.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append_series_as_row.png">
<p><span class="yiyi-st" id="yiyi-153">您应该使用<code class="docutils literal"><span class="pre">ignore_index</span></code>与此方法指示DataFrame丢弃其索引。</span><span class="yiyi-st" id="yiyi-154">如果希望保留索引，应该构造一个适当索引的DataFrame，并附加或连接这些对象。</span></p>
<p><span class="yiyi-st" id="yiyi-155">您还可以传递一个列表或系列：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [36]: </span><span class="n">dicts</span> <span class="o">=</span> <span class="p">[{</span><span class="s1">'A'</span><span class="p">:</span> <span class="mi">1</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">:</span> <span class="mi">2</span><span class="p">,</span> <span class="s1">'C'</span><span class="p">:</span> <span class="mi">3</span><span class="p">,</span> <span class="s1">'X'</span><span class="p">:</span> <span class="mi">4</span><span class="p">},</span>
<span class="gp"> ....:</span> <span class="p">{</span><span class="s1">'A'</span><span class="p">:</span> <span class="mi">5</span><span class="p">,</span> <span class="s1">'B'</span><span class="p">:</span> <span class="mi">6</span><span class="p">,</span> <span class="s1">'C'</span><span class="p">:</span> <span class="mi">7</span><span class="p">,</span> <span class="s1">'Y'</span><span class="p">:</span> <span class="mi">8</span><span class="p">}]</span>
<span class="gp"> ....:</span>
<span class="gp">In [37]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">dicts</span><span class="p">,</span> <span class="n">ignore_index</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append_dits.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_append_dits.png">
</div>
</div>
<div class="section" id="database-style-dataframe-joining-merging">
<span id="merging-join"></span><h2><span class="yiyi-st" id="yiyi-156">Database-style DataFrame joining/merging</span></h2>
<p><span class="yiyi-st" id="yiyi-157">pandas具有全功能的，<strong>高性能</strong>内存中连接操作，与SQL等关系数据库非常相似。</span><span class="yiyi-st" id="yiyi-158">这些方法比其他开源实现（例如R中的<code class="docutils literal"><span class="pre">base::merge.data.frame</span></code>）执行得更好（在某些情况下好得多一个数量级）。</span><span class="yiyi-st" id="yiyi-159">其原因是DataFrame中的数据的仔细的算法设计和内部布局。</span></p>
<p><span class="yiyi-st" id="yiyi-160">有关某些高级策略，请参阅<a class="reference internal" href="cookbook.html#cookbook-merge"><span class="std std-ref">cookbook</span></a>。</span></p>
<p><span class="yiyi-st" id="yiyi-161">熟悉SQL但是新增了pandas的用户可能对与SQL的<a class="reference internal" href="comparison_with_sql.html#compare-with-sql-join"><span class="std std-ref">comparison with SQL</span></a></span></p>
<p><span class="yiyi-st" id="yiyi-162">pandas提供单个函数<code class="docutils literal"><span class="pre">merge</span></code>作为DataFrame对象之间的所有标准数据库连接操作的入口点：</span></p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">left_on</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span> <span class="n">right_on</span><span class="o">=</span><span class="kc">None</span><span class="p">,</span>
<span class="n">left_index</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="n">right_index</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="n">sort</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span>
<span class="n">suffixes</span><span class="o">=</span><span class="p">(</span><span class="s1">'_x'</span><span class="p">,</span> <span class="s1">'_y'</span><span class="p">),</span> <span class="n">copy</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">indicator</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
</pre></div>
</div>
<ul>
<li><p class="first"><span class="yiyi-st" id="yiyi-163"><code class="docutils literal"><span class="pre">left</span></code>：DataFrame对象</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-164"><code class="docutils literal"><span class="pre">right</span></code>：另一个DataFrame对象</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-165"><code class="docutils literal"><span class="pre">on</span></code>：要加入的列（名称）。</span><span class="yiyi-st" id="yiyi-166">必须在左右DataFrame对象中找到。</span><span class="yiyi-st" id="yiyi-167">如果未传递，且<code class="docutils literal"><span class="pre">left_index</span></code>和<code class="docutils literal"><span class="pre">right_index</span></code>为<code class="docutils literal"><span class="pre">False</span></code>，则DataFrames中的列的交集将被推断为连接键</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-168"><code class="docutils literal"><span class="pre">left_on</span></code>：左侧DataFrame中用作键的列。</span><span class="yiyi-st" id="yiyi-169">可以是列名称或长度等于DataFrame长度的数组</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-170"><code class="docutils literal"><span class="pre">right_on</span></code>：来自右侧DataFrame的列，用作键。</span><span class="yiyi-st" id="yiyi-171">可以是列名称或长度等于DataFrame长度的数组</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-172"><code class="docutils literal"><span class="pre">left_index</span></code>：如果<code class="docutils literal"><span class="pre">True</span></code>，请使用左侧DataFrame中的索引（行标签）作为其连接键。</span><span class="yiyi-st" id="yiyi-173">在具有MultiIndex（分层）的DataFrame的情况下，级别数必须与来自右侧DataFrame的连接键数匹配</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-174"><code class="docutils literal"><span class="pre">right_index</span></code>：与<code class="docutils literal"><span class="pre">left_index</span></code>使用方式相同，适用于正确的DataFrame</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-175"><code class="docutils literal"><span class="pre">how</span></code>：<code class="docutils literal"><span class="pre">'left'</span></code>，<code class="docutils literal"><span class="pre">'right'</span></code>，<code class="docutils literal"><span class="pre">'outer'</span></code>，<code class="docutils literal"><span class="pre">'inner'</span></code>。</span><span class="yiyi-st" id="yiyi-176">默认为<code class="docutils literal"><span class="pre">inner</span></code>。</span><span class="yiyi-st" id="yiyi-177">有关每种方法的详细说明，请参阅下文</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-178"><code class="docutils literal"><span class="pre">sort</span></code>：按照字典顺序通过连接键对结果DataFrame进行排序。</span><span class="yiyi-st" id="yiyi-179">默认为<code class="docutils literal"><span class="pre">True</span></code>，设置为<code class="docutils literal"><span class="pre">False</span></code>会大幅提高性能</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-180"><code class="docutils literal"><span class="pre">suffixes</span></code>：应用于重叠列的字符串后缀的元组。</span><span class="yiyi-st" id="yiyi-181">默认为<code class="docutils literal"><span class="pre">（'_ x'，</span> <span class="pre">'_ y'）</span></code>。</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-182"><code class="docutils literal"><span class="pre">copy</span></code>：始终从传递的DataFrame对象复制数据（默认<code class="docutils literal"><span class="pre">True</span></code>），即使不需要重建索引。</span><span class="yiyi-st" id="yiyi-183">在许多情况下不能避免，但可以提高性能/内存使用。</span><span class="yiyi-st" id="yiyi-184">可以避免复制的情况有些病态，但仍然提供此选项。</span></p>
</li>
<li><p class="first"><span class="yiyi-st" id="yiyi-185"><code class="docutils literal"><span class="pre">indicator</span></code>：向输出DataFrame中添加一个名为<code class="docutils literal"><span class="pre">_merge</span></code>的列，其中包含有关每行源的信息。</span><span class="yiyi-st" id="yiyi-186"><code class="docutils literal"><span class="pre">_merge</span></code> is Categorical-type and takes on a value of <code class="docutils literal"><span class="pre">left_only</span></code> for observations whose merge key only appears in <code class="docutils literal"><span class="pre">'left'</span></code> DataFrame, <code class="docutils literal"><span class="pre">right_only</span></code> for observations whose merge key only appears in <code class="docutils literal"><span class="pre">'right'</span></code> DataFrame, and <code class="docutils literal"><span class="pre">both</span></code> if the observation’s merge key is found in both.</span></p>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-187"><span class="versionmodified">版本0.17.0中的新功能。</span></span></p>
</div>
</li>
</ul>
<p><span class="yiyi-st" id="yiyi-188">返回类型将与<code class="docutils literal"><span class="pre">left</span></code></span><span class="yiyi-st" id="yiyi-189">如果<code class="docutils literal"><span class="pre">left</span></code>是<code class="docutils literal"><span class="pre">DataFrame</span></code>和<code class="docutils literal"><span class="pre">right</span></code>是DataFrame的子类，则返回类型仍然是<code class="docutils literal"><span class="pre">DataFrame</span></code>。</span></p>
<p><span class="yiyi-st" id="yiyi-190"><code class="docutils literal"><span class="pre">merge</span></code>是pandas命名空间中的函数，它也可用作DataFrame实例方法，调用DataFrame被隐式地视为连接中的左侧对象。</span></p>
<p><span class="yiyi-st" id="yiyi-191">The related <code class="docutils literal"><span class="pre">DataFrame.join</span></code> method, uses <code class="docutils literal"><span class="pre">merge</span></code> internally for the index-on-index (by default) and column(s)-on-index join. </span><span class="yiyi-st" id="yiyi-192">如果您只加入索引，您可能希望使用<code class="docutils literal"><span class="pre">DataFrame.join</span></code>来保存自己一些输入。</span></p>
<div class="section" id="brief-primer-on-merge-methods-relational-algebra">
<h3><span class="yiyi-st" id="yiyi-193">Brief primer on merge methods (relational algebra)</span></h3>
<p><span class="yiyi-st" id="yiyi-194">经验丰富的关系数据库（如SQL）的用户将熟悉用于描述两个类似SQL表结构（DataFrame对象）之间的连接操作的术语。</span><span class="yiyi-st" id="yiyi-195">有几种情况需要考虑，这是非常重要的理解：</span></p>
<ul class="simple">
<li><span class="yiyi-st" id="yiyi-196"><strong>一对一</strong>连接：例如，当在其索引（必须包含唯一值）上连接两个DataFrame对象时，</span></li>
<li><span class="yiyi-st" id="yiyi-197"><strong>多对一</strong>连接：例如，当将索引（唯一）连接到DataFrame中的一个或多个列时</span></li>
<li><span class="yiyi-st" id="yiyi-198"><strong>多对多</strong>连接：连接列上的列。</span></li>
</ul>
<div class="admonition note">
<p class="first admonition-title"><span class="yiyi-st" id="yiyi-199">注意</span></p>
<p class="last"><span class="yiyi-st" id="yiyi-200">当连接列上的列（可能是多对多连接）时，传递的DataFrame对象<strong>上的任何索引都将被丢弃</strong>。</span></p>
</div>
<p><span class="yiyi-st" id="yiyi-201">值得花一些时间来理解<strong>多对多</strong>连接情况的结果。</span><span class="yiyi-st" id="yiyi-202">在SQL /标准关系代数中，如果一个键组合在两个表中出现多次，则结果表将具有相关数据的<strong>笛卡尔乘积</strong>。</span><span class="yiyi-st" id="yiyi-203">这里是一个非常基本的例子，一个唯一的组合键：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [38]: </span><span class="n">left</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="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'K3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'A'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</span><span class="p">,</span> <span class="s1">'A3'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">,</span> <span class="s1">'B3'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [39]: </span><span class="n">right</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="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'K3'</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">'C0'</span><span class="p">,</span> <span class="s1">'C1'</span><span class="p">,</span> <span class="s1">'C2'</span><span class="p">,</span> <span class="s1">'C3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D1'</span><span class="p">,</span> <span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [40]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="s1">'key'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key.png">
<p><span class="yiyi-st" id="yiyi-204">这里是一个更复杂的示例与多个连接键：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [41]: </span><span class="n">left</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">'key1'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'key2'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'A'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</span><span class="p">,</span> <span class="s1">'A3'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">,</span> <span class="s1">'B3'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [42]: </span><span class="n">right</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">'key1'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'key2'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</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">'C0'</span><span class="p">,</span> <span class="s1">'C1'</span><span class="p">,</span> <span class="s1">'C2'</span><span class="p">,</span> <span class="s1">'C3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D1'</span><span class="p">,</span> <span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [43]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_multiple.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_multiple.png">
<p><span class="yiyi-st" id="yiyi-205">The <code class="docutils literal"><span class="pre">how</span></code> argument to <code class="docutils literal"><span class="pre">merge</span></code> specifies how to determine which keys are to be included in the resulting table. </span><span class="yiyi-st" id="yiyi-206">如果在左或右表中未出现<strong>组合键</strong>，则连接表中的值将为<code class="docutils literal"><span class="pre">NA</span></code>。</span><span class="yiyi-st" id="yiyi-207">以下是<code class="docutils literal"><span class="pre">how</span></code>选项及其SQL等效名称的摘要：</span></p>
<table border="1" class="docutils">
<colgroup>
<col width="20%">
<col width="20%">
<col width="60%">
</colgroup>
<thead valign="bottom">
<tr class="row-odd"><th class="head"><span class="yiyi-st" id="yiyi-208">合并方法</span></th>
<th class="head"><span class="yiyi-st" id="yiyi-209">SQL加入名称</span></th>
<th class="head"><span class="yiyi-st" id="yiyi-210">描述</span></th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even"><td><span class="yiyi-st" id="yiyi-211"><code class="docutils literal"><span class="pre">left</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-212"><code class="docutils literal"><span class="pre">LEFT</span> <span class="pre">OUTER</span> <span class="pre">JOIN</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-213">仅使用左框架的键</span></td>
</tr>
<tr class="row-odd"><td><span class="yiyi-st" id="yiyi-214"><code class="docutils literal"><span class="pre">right</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-215"><code class="docutils literal"><span class="pre">RIGHT</span> <span class="pre">OUTER</span> <span class="pre">JOIN</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-216">仅使用右边框的键</span></td>
</tr>
<tr class="row-even"><td><span class="yiyi-st" id="yiyi-217"><code class="docutils literal"><span class="pre">outer</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-218"><code class="docutils literal"><span class="pre">FULL</span> <span class="pre">OUTER</span> <span class="pre">JOIN</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-219">使用来自两个帧的键的联合</span></td>
</tr>
<tr class="row-odd"><td><span class="yiyi-st" id="yiyi-220"><code class="docutils literal"><span class="pre">inner</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-221"><code class="docutils literal"><span class="pre">INNER</span> <span class="pre">JOIN</span></code></span></td>
<td><span class="yiyi-st" id="yiyi-222">使用两个帧的交叉点</span></td>
</tr>
</tbody>
</table>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [44]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'left'</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_left.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_left.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [45]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'right'</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_right.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_right.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [46]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'outer'</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_outer.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_outer.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [47]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_inner.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_on_key_inner.png">
</div>
<div class="section" id="the-merge-indicator">
<span id="merging-indicator"></span><h3><span class="yiyi-st" id="yiyi-223">The merge indicator</span></h3>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-224"><span class="versionmodified">版本0.17.0中的新功能。</span></span></p>
</div>
<p><span class="yiyi-st" id="yiyi-225"><code class="docutils literal"><span class="pre">merge</span></code>现在接受参数<code class="docutils literal"><span class="pre">indicator</span></code>。</span><span class="yiyi-st" id="yiyi-226">如果<code class="docutils literal"><span class="pre">True</span></code>，则将一个名为<code class="docutils literal"><span class="pre">_merge</span></code>的分类类型列添加到接受值的输出对象：</span></p>
<blockquote>
<div><table border="1" class="docutils">
<colgroup>
<col width="69%">
<col width="31%">
</colgroup>
<thead valign="bottom">
<tr class="row-odd"><th class="head"><span class="yiyi-st" id="yiyi-227">观察原产地</span></th>
<th class="head"><span class="yiyi-st" id="yiyi-228"><code class="docutils literal"><span class="pre">_merge</span></code>值</span></th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even"><td><span class="yiyi-st" id="yiyi-229">只在<code class="docutils literal"><span class="pre">'left'</span></code>框中合并键</span></td>
<td><span class="yiyi-st" id="yiyi-230"><code class="docutils literal"><span class="pre">left_only</span></code></span></td>
</tr>
<tr class="row-odd"><td><span class="yiyi-st" id="yiyi-231">仅在<code class="docutils literal"><span class="pre">'right'</span></code>框中合并键</span></td>
<td><span class="yiyi-st" id="yiyi-232"><code class="docutils literal"><span class="pre">right_only</span></code></span></td>
</tr>
<tr class="row-even"><td><span class="yiyi-st" id="yiyi-233">在两个框架中合并关键帧</span></td>
<td><span class="yiyi-st" id="yiyi-234"><code class="docutils literal"><span class="pre">both</span></code></span></td>
</tr>
</tbody>
</table>
</div></blockquote>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [48]: </span><span class="n">df1</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">'col1'</span><span class="p">:</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">],</span> <span class="s1">'col_left'</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 [49]: </span><span class="n">df2</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">'col1'</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="s1">'col_right'</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="gp">In [50]: </span><span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">df1</span><span class="p">,</span> <span class="n">df2</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="s1">'col1'</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'outer'</span><span class="p">,</span> <span class="n">indicator</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="gr">Out[50]: </span>
<span class="go"> col1 col_left col_right _merge</span>
<span class="go">0 0 a NaN left_only</span>
<span class="go">1 1 b 2.0 both</span>
<span class="go">2 2 NaN 2.0 right_only</span>
<span class="go">3 2 NaN 2.0 right_only</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-235"><code class="docutils literal"><span class="pre">indicator</span></code>参数也将接受字符串参数，在这种情况下，指示符函数将使用传递的字符串的值作为指示符列的名称。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [51]: </span><span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">df1</span><span class="p">,</span> <span class="n">df2</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="s1">'col1'</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'outer'</span><span class="p">,</span> <span class="n">indicator</span><span class="o">=</span><span class="s1">'indicator_column'</span><span class="p">)</span>
<span class="gr">Out[51]: </span>
<span class="go"> col1 col_left col_right indicator_column</span>
<span class="go">0 0 a NaN left_only</span>
<span class="go">1 1 b 2.0 both</span>
<span class="go">2 2 NaN 2.0 right_only</span>
<span class="go">3 2 NaN 2.0 right_only</span>
</pre></div>
</div>
</div>
<div class="section" id="joining-on-index">
<span id="merging-join-index"></span><h3><span class="yiyi-st" id="yiyi-236">Joining on index</span></h3>
<p><span class="yiyi-st" id="yiyi-237"><code class="docutils literal"><span class="pre">DataFrame.join</span></code>是一种方便的方法，用于将两个可能不同索引的DataFrames的列合并为单个结果DataFrame。</span><span class="yiyi-st" id="yiyi-238">这里有一个非常基本的例子：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [52]: </span><span class="n">left</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">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [53]: </span><span class="n">right</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">'C'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'C0'</span><span class="p">,</span> <span class="s1">'C2'</span><span class="p">,</span> <span class="s1">'C3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'K3'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [54]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [55]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'outer'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_outer.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_outer.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [56]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_inner.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_inner.png">
<p><span class="yiyi-st" id="yiyi-239">这里的数据对齐在索引（行标签）上。</span><span class="yiyi-st" id="yiyi-240">使用<code class="docutils literal"><span class="pre">merge</span></code>加上指示它使用索引的其他参数也可以实现相同的行为：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [57]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">left_index</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span> <span class="n">right_index</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'outer'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_index_outer.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_index_outer.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [58]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">left_index</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span> <span class="n">right_index</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">);</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_index_inner.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_index_inner.png">
</div>
<div class="section" id="joining-key-columns-on-an-index">
<h3><span class="yiyi-st" id="yiyi-241">Joining key columns on an index</span></h3>
<p><span class="yiyi-st" id="yiyi-242"><code class="docutils literal"><span class="pre">join</span></code>在参数上接受一个可选的<code class="docutils literal"><span class="pre">on</span></code></span><span class="yiyi-st" id="yiyi-243">这两个函数调用是完全等价的：</span></p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="n">key_or_keys</span><span class="p">)</span>
<span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">left_on</span><span class="o">=</span><span class="n">key_or_keys</span><span class="p">,</span> <span class="n">right_index</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span>
<span class="n">how</span><span class="o">=</span><span class="s1">'left'</span><span class="p">,</span> <span class="n">sort</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-244">显然你可以选择任何形式，你觉得更方便。</span><span class="yiyi-st" id="yiyi-245">对于多对一连接（其中一个DataFrame已通过连接键索引），使用<code class="docutils literal"><span class="pre">join</span></code>可能更方便。</span><span class="yiyi-st" id="yiyi-246">这里有一个简单的例子：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [59]: </span><span class="n">left</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">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</span><span class="p">,</span> <span class="s1">'A3'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">,</span> <span class="s1">'B3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'key'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [60]: </span><span class="n">right</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">'C'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'C0'</span><span class="p">,</span> <span class="s1">'C1'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D1'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [61]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="s1">'key'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_key_columns.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_key_columns.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [62]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">left_on</span><span class="o">=</span><span class="s1">'key'</span><span class="p">,</span> <span class="n">right_index</span><span class="o">=</span><span class="bp">True</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">how</span><span class="o">=</span><span class="s1">'left'</span><span class="p">,</span> <span class="n">sort</span><span class="o">=</span><span class="bp">False</span><span class="p">);</span>
<span class="gp"> ....:</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_key_columns.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_key_columns.png">
<p id="merging-multikey-join"><span class="yiyi-st" id="yiyi-247">要在多个键上连接，传递的DataFrame必须具有<code class="docutils literal"><span class="pre">MultiIndex</span></code>：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [63]: </span><span class="n">left</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">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</span><span class="p">,</span> <span class="s1">'A3'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">,</span> <span class="s1">'B3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'key1'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'key2'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [64]: </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="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">)])</span>
<span class="gp"> ....:</span>
<span class="gp">In [65]: </span><span class="n">right</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">'C'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'C0'</span><span class="p">,</span> <span class="s1">'C1'</span><span class="p">,</span> <span class="s1">'C2'</span><span class="p">,</span> <span class="s1">'C3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D1'</span><span class="p">,</span> <span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="n">index</span><span class="p">)</span>
<span class="gp"> ....:</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-248">现在可以通过传递两个键列名称来连接：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [66]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multikeys.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multikeys.png">
<p id="merging-df-inner-join"><span class="yiyi-st" id="yiyi-249"><code class="docutils literal"><span class="pre">DataFrame.join</span></code>的默认值是执行左连接（本质上是一个“VLOOKUP”操作，对于Excel用户），它只使用在调用DataFrame中找到的键。</span><span class="yiyi-st" id="yiyi-250">其他连接类型，例如内连接，可以很容易地执行：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [67]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key1'</span><span class="p">,</span> <span class="s1">'key2'</span><span class="p">],</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multikeys_inner.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multikeys_inner.png">
<p><span class="yiyi-st" id="yiyi-251">正如你所看到的，这会删除任何没有匹配的行。</span></p>
</div>
<div class="section" id="joining-a-single-index-to-a-multi-index">
<span id="merging-join-on-mi"></span><h3><span class="yiyi-st" id="yiyi-252">Joining a single Index to a Multi-index</span></h3>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-253"><span class="versionmodified">版本0.14.0中的新功能。</span></span></p>
</div>
<p><span class="yiyi-st" id="yiyi-254">您可以使用多索引的<code class="docutils literal"><span class="pre">DataFrame</span></code>级别加入单索引的<code class="docutils literal"><span class="pre">DataFrame</span></code>。</span><span class="yiyi-st" id="yiyi-255">该级别将使单索引帧的索引的名称与多索引帧的级别名称匹配。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [68]: </span><span class="n">left</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">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="n">pd</span><span class="o">.</span><span class="n">Index</span><span class="p">([</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">],</span> <span class="n">name</span><span class="o">=</span><span class="s1">'key'</span><span class="p">))</span>
<span class="gp"> ....:</span>
<span class="gp">In [69]: </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="s1">'K0'</span><span class="p">,</span> <span class="s1">'Y0'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'Y1'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'Y2'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'Y3'</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">'key'</span><span class="p">,</span> <span class="s1">'Y'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [70]: </span><span class="n">right</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">'C'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'C0'</span><span class="p">,</span> <span class="s1">'C1'</span><span class="p">,</span> <span class="s1">'C2'</span><span class="p">,</span> <span class="s1">'C3'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'D'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'D0'</span><span class="p">,</span> <span class="s1">'D1'</span><span class="p">,</span> <span class="s1">'D2'</span><span class="p">,</span> <span class="s1">'D3'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="n">index</span><span class="p">)</span>
<span class="gp"> ....:</span>
<span class="gp">In [71]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multiindex_inner.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multiindex_inner.png">
<p><span class="yiyi-st" id="yiyi-256">这是等效的，但是较少冗长和更多的内存高效/更快。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [72]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="o">.</span><span class="n">reset_index</span><span class="p">(),</span> <span class="n">right</span><span class="o">.</span><span class="n">reset_index</span><span class="p">(),</span>
<span class="gp"> ....:</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key'</span><span class="p">],</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">)</span><span class="o">.</span><span class="n">set_index</span><span class="p">([</span><span class="s1">'key'</span><span class="p">,</span><span class="s1">'Y'</span><span class="p">])</span>
<span class="gp"> ....:</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_multiindex_alternative.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_multiindex_alternative.png">
</div>
<div class="section" id="joining-with-two-multi-indexes">
<h3><span class="yiyi-st" id="yiyi-257">Joining with two multi-indexes</span></h3>
<p><span class="yiyi-st" id="yiyi-258">这不是通过<code class="docutils literal"><span class="pre">join</span></code>实现的，但是可以使用以下方法完成。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [73]: </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="s1">'K0'</span><span class="p">,</span> <span class="s1">'X0'</span><span class="p">),</span> <span class="p">(</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'X1'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="p">(</span><span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'X2'</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">'key'</span><span class="p">,</span> <span class="s1">'X'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [74]: </span><span class="n">left</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">'A0'</span><span class="p">,</span> <span class="s1">'A1'</span><span class="p">,</span> <span class="s1">'A2'</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">'B0'</span><span class="p">,</span> <span class="s1">'B1'</span><span class="p">,</span> <span class="s1">'B2'</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">index</span><span class="o">=</span><span class="n">index</span><span class="p">)</span>
<span class="gp"> ....:</span>
<span class="gp">In [75]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="o">.</span><span class="n">reset_index</span><span class="p">(),</span> <span class="n">right</span><span class="o">.</span><span class="n">reset_index</span><span class="p">(),</span>
<span class="gp"> ....:</span> <span class="n">on</span><span class="o">=</span><span class="p">[</span><span class="s1">'key'</span><span class="p">],</span> <span class="n">how</span><span class="o">=</span><span class="s1">'inner'</span><span class="p">)</span><span class="o">.</span><span class="n">set_index</span><span class="p">([</span><span class="s1">'key'</span><span class="p">,</span><span class="s1">'X'</span><span class="p">,</span><span class="s1">'Y'</span><span class="p">])</span>
<span class="gp"> ....:</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_two_multiindex.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_two_multiindex.png">
</div>
<div class="section" id="overlapping-value-columns">
<h3><span class="yiyi-st" id="yiyi-259">Overlapping value columns</span></h3>
<p><span class="yiyi-st" id="yiyi-260">合并<code class="docutils literal"><span class="pre">suffixes</span></code>参数需要一个字符串列表的元组，以附加到输入DataFrames中的重叠列名称，以消除结果列的歧义：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [76]: </span><span class="n">left</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">'k'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">],</span> <span class="s1">'v'</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">In [77]: </span><span class="n">right</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">'k'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K3'</span><span class="p">],</span> <span class="s1">'v'</span><span class="p">:</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">6</span><span class="p">]})</span>
<span class="gp">In [78]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="s1">'k'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_overlapped.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_overlapped.png">
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [79]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">merge</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">on</span><span class="o">=</span><span class="s1">'k'</span><span class="p">,</span> <span class="n">suffixes</span><span class="o">=</span><span class="p">[</span><span class="s1">'_l'</span><span class="p">,</span> <span class="s1">'_r'</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_overlapped_suffix.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_overlapped_suffix.png">
<p><span class="yiyi-st" id="yiyi-261"><code class="docutils literal"><span class="pre">DataFrame.join</span></code>具有类似的<code class="docutils literal"><span class="pre">lsuffix</span></code>和<code class="docutils literal"><span class="pre">rsuffix</span></code>参数。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [80]: </span><span class="n">left</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">set_index</span><span class="p">(</span><span class="s1">'k'</span><span class="p">)</span>
<span class="gp">In [81]: </span><span class="n">right</span> <span class="o">=</span> <span class="n">right</span><span class="o">.</span><span class="n">set_index</span><span class="p">(</span><span class="s1">'k'</span><span class="p">)</span>
<span class="gp">In [82]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">right</span><span class="p">,</span> <span class="n">lsuffix</span><span class="o">=</span><span class="s1">'_l'</span><span class="p">,</span> <span class="n">rsuffix</span><span class="o">=</span><span class="s1">'_r'</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_overlapped_multi_suffix.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_merge_overlapped_multi_suffix.png">
</div>
<div class="section" id="joining-multiple-dataframe-or-panel-objects">
<span id="merging-multiple-join"></span><h3><span class="yiyi-st" id="yiyi-262">Joining multiple DataFrame or Panel objects</span></h3>
<p><span class="yiyi-st" id="yiyi-263">DataFrames的列表或元组也可以传递到<code class="docutils literal"><span class="pre">DataFrame.join</span></code>，以将它们的索引连接在一起。</span><span class="yiyi-st" id="yiyi-264">对于<code class="docutils literal"><span class="pre">Panel.join</span></code>也是如此。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [83]: </span><span class="n">right2</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">'v'</span><span class="p">:</span> <span class="p">[</span><span class="mi">7</span><span class="p">,</span> <span class="mi">8</span><span class="p">,</span> <span class="mi">9</span><span class="p">]},</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">])</span>
<span class="gp">In [84]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">left</span><span class="o">.</span><span class="n">join</span><span class="p">([</span><span class="n">right</span><span class="p">,</span> <span class="n">right2</span><span class="p">])</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multi_df.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_join_multi_df.png">
</div>
<div class="section" id="merging-together-values-within-series-or-dataframe-columns">
<span id="merging-combine-first-update"></span><h3><span class="yiyi-st" id="yiyi-265">Merging together values within Series or DataFrame columns</span></h3>
<p><span class="yiyi-st" id="yiyi-266">另一个相当普遍的情况是有两个相似索引（或类似索引）的Series或DataFrame对象，并且希望在一个对象中“修补”值，以匹配另一个中的索引值。</span><span class="yiyi-st" id="yiyi-267">这里是一个例子：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [85]: </span><span class="n">df1</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">nan</span><span class="p">,</span> <span class="mf">3.</span><span class="p">,</span> <span class="mf">5.</span><span class="p">],</span> <span class="p">[</span><span class="o">-</span><span class="mf">4.6</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="n">np</span><span class="o">.</span><span class="n">nan</span><span class="p">],</span>
<span class="gp"> ....:</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="mf">7.</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="gp"> ....:</span>
<span class="gp">In [86]: </span><span class="n">df2</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="o">-</span><span class="mf">42.6</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="o">-</span><span class="mf">8.2</span><span class="p">],</span> <span class="p">[</span><span class="o">-</span><span class="mf">5.</span><span class="p">,</span> <span class="mf">1.6</span><span class="p">,</span> <span class="mi">4</span><span class="p">]],</span>
<span class="gp"> ....:</span> <span class="n">index</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="gp"> ....:</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-268">为此，请使用<code class="docutils literal"><span class="pre">combine_first</span></code>方法：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [87]: </span><span class="n">result</span> <span class="o">=</span> <span class="n">df1</span><span class="o">.</span><span class="n">combine_first</span><span class="p">(</span><span class="n">df2</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_combine_first.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_combine_first.png">
<p><span class="yiyi-st" id="yiyi-269">注意，这个方法只从右边的DataFrame中获取值，如果它们在左边的DataFrame中缺失的话。</span><span class="yiyi-st" id="yiyi-270">相关方法<code class="docutils literal"><span class="pre">update</span></code>可替代地更改非NA值：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [88]: </span><span class="n">df1</span><span class="o">.</span><span class="n">update</span><span class="p">(</span><span class="n">df2</span><span class="p">)</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span>
</pre></div>
</div>
<img alt="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_update.png" src="http://pandas.pydata.org/pandas-docs/version/0.19.2/_images/merging_update.png">
</div>
</div>
<div class="section" id="timeseries-friendly-merging">
<span id="merging-time-series"></span><h2><span class="yiyi-st" id="yiyi-271">Timeseries friendly merging</span></h2>
<div class="section" id="merging-ordered-data">
<span id="merging-merge-ordered"></span><h3><span class="yiyi-st" id="yiyi-272">Merging Ordered Data</span></h3>
<p><span class="yiyi-st" id="yiyi-273"><a class="reference internal" href="generated/pandas.merge_ordered.html#pandas.merge_ordered" title="pandas.merge_ordered"><code class="xref py py-func docutils literal"><span class="pre">merge_ordered()</span></code></a>函数允许组合时间序列和其他有序数据。</span><span class="yiyi-st" id="yiyi-274">特别地，它具有可选的<code class="docutils literal"><span class="pre">fill_method</span></code>关键字来填充/内插缺失的数据：</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [89]: </span><span class="n">left</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">'k'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K0'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'lv'</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="mi">4</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'s'</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="s1">'d'</span><span class="p">]})</span>
<span class="gp"> ....:</span>
<span class="gp">In [90]: </span><span class="n">right</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">'k'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'K1'</span><span class="p">,</span> <span class="s1">'K2'</span><span class="p">,</span> <span class="s1">'K4'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'rv'</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 [91]: </span><span class="n">pd</span><span class="o">.</span><span class="n">merge_ordered</span><span class="p">(</span><span class="n">left</span><span class="p">,</span> <span class="n">right</span><span class="p">,</span> <span class="n">fill_method</span><span class="o">=</span><span class="s1">'ffill'</span><span class="p">,</span> <span class="n">left_by</span><span class="o">=</span><span class="s1">'s'</span><span class="p">)</span>
<span class="gr">Out[91]: </span>
<span class="go"> k lv s rv</span>
<span class="go">0 K0 1.0 a NaN</span>
<span class="go">1 K1 1.0 a 1.0</span>
<span class="go">2 K2 1.0 a 2.0</span>
<span class="go">3 K4 1.0 a 3.0</span>
<span class="go">4 K1 2.0 b 1.0</span>
<span class="go">5 K2 2.0 b 2.0</span>
<span class="go">6 K4 2.0 b 3.0</span>
<span class="go">7 K1 3.0 c 1.0</span>
<span class="go">8 K2 3.0 c 2.0</span>
<span class="go">9 K4 3.0 c 3.0</span>
<span class="go">10 K1 NaN d 1.0</span>
<span class="go">11 K2 4.0 d 2.0</span>
<span class="go">12 K4 4.0 d 3.0</span>
</pre></div>
</div>
</div>
<div class="section" id="merging-asof">
<span id="merging-merge-asof"></span><h3><span class="yiyi-st" id="yiyi-275">Merging AsOf</span></h3>
<div class="versionadded">
<p><span class="yiyi-st" id="yiyi-276"><span class="versionmodified">版本0.19.0中的新功能。</span></span></p>
</div>
<p><span class="yiyi-st" id="yiyi-277"><a class="reference internal" href="generated/pandas.merge_asof.html#pandas.merge_asof" title="pandas.merge_asof"><code class="xref py py-func docutils literal"><span class="pre">merge_asof()</span></code></a>类似于有序左连接，除了我们匹配最近的键而不是相等的键。</span><span class="yiyi-st" id="yiyi-278">For each row in the <code class="docutils literal"><span class="pre">left</span></code> DataFrame, we select the last row in the <code class="docutils literal"><span class="pre">right</span></code> DataFrame whose <code class="docutils literal"><span class="pre">on</span></code> key is less than the left’s key. </span><span class="yiyi-st" id="yiyi-279">两个DataFrames都必须按键排序。</span></p>
<p><span class="yiyi-st" id="yiyi-280">可选地，asof合并可以执行分组合并。</span><span class="yiyi-st" id="yiyi-281">除了<code class="docutils literal"><span class="pre">on</span></code>键上最接近的匹配，这与<code class="docutils literal"><span class="pre">by</span></code>键相同。</span></p>
<p><span class="yiyi-st" id="yiyi-282">例如；我们可能会有<code class="docutils literal"><span class="pre">trades</span></code>和<code class="docutils literal"><span class="pre">quotes</span></code>，我们要<code class="docutils literal"><span class="pre">asof</span></code>合并它们。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [92]: </span><span class="n">trades</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="gp"> ....:</span> <span class="s1">'time'</span><span class="p">:</span> <span class="n">pd</span><span class="o">.</span><span class="n">to_datetime</span><span class="p">([</span><span class="s1">'20160525 13:30:00.023'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.038'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.048'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.048'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.048'</span><span class="p">]),</span>
<span class="gp"> ....:</span> <span class="s1">'ticker'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'MSFT'</span><span class="p">,</span> <span class="s1">'MSFT'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'GOOG'</span><span class="p">,</span> <span class="s1">'GOOG'</span><span class="p">,</span> <span class="s1">'AAPL'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'price'</span><span class="p">:</span> <span class="p">[</span><span class="mf">51.95</span><span class="p">,</span> <span class="mf">51.95</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="mf">720.77</span><span class="p">,</span> <span class="mf">720.92</span><span class="p">,</span> <span class="mf">98.00</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'quantity'</span><span class="p">:</span> <span class="p">[</span><span class="mi">75</span><span class="p">,</span> <span class="mi">155</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="mi">100</span><span class="p">,</span> <span class="mi">100</span><span class="p">,</span> <span class="mi">100</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'time'</span><span class="p">,</span> <span class="s1">'ticker'</span><span class="p">,</span> <span class="s1">'price'</span><span class="p">,</span> <span class="s1">'quantity'</span><span class="p">])</span>
<span class="gp"> ....:</span>
<span class="gp">In [93]: </span><span class="n">quotes</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="gp"> ....:</span> <span class="s1">'time'</span><span class="p">:</span> <span class="n">pd</span><span class="o">.</span><span class="n">to_datetime</span><span class="p">([</span><span class="s1">'20160525 13:30:00.023'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.023'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.030'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.041'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.048'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.049'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.072'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'20160525 13:30:00.075'</span><span class="p">]),</span>
<span class="gp"> ....:</span> <span class="s1">'ticker'</span><span class="p">:</span> <span class="p">[</span><span class="s1">'GOOG'</span><span class="p">,</span> <span class="s1">'MSFT'</span><span class="p">,</span> <span class="s1">'MSFT'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'MSFT'</span><span class="p">,</span> <span class="s1">'GOOG'</span><span class="p">,</span> <span class="s1">'AAPL'</span><span class="p">,</span> <span class="s1">'GOOG'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="s1">'MSFT'</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'bid'</span><span class="p">:</span> <span class="p">[</span><span class="mf">720.50</span><span class="p">,</span> <span class="mf">51.95</span><span class="p">,</span> <span class="mf">51.97</span><span class="p">,</span> <span class="mf">51.99</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="mf">720.50</span><span class="p">,</span> <span class="mf">97.99</span><span class="p">,</span> <span class="mf">720.50</span><span class="p">,</span> <span class="mf">52.01</span><span class="p">],</span>
<span class="gp"> ....:</span> <span class="s1">'ask'</span><span class="p">:</span> <span class="p">[</span><span class="mf">720.93</span><span class="p">,</span> <span class="mf">51.96</span><span class="p">,</span> <span class="mf">51.98</span><span class="p">,</span> <span class="mf">52.00</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="mf">720.93</span><span class="p">,</span> <span class="mf">98.01</span><span class="p">,</span> <span class="mf">720.88</span><span class="p">,</span> <span class="mf">52.03</span><span class="p">]},</span>
<span class="gp"> ....:</span> <span class="n">columns</span><span class="o">=</span><span class="p">[</span><span class="s1">'time'</span><span class="p">,</span> <span class="s1">'ticker'</span><span class="p">,</span> <span class="s1">'bid'</span><span class="p">,</span> <span class="s1">'ask'</span><span class="p">])</span>
<span class="gp"> ....:</span>
</pre></div>
</div>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [94]: </span><span class="n">trades</span>
<span class="gr">Out[94]: </span>
<span class="go"> time ticker price quantity</span>
<span class="go">0 2016-05-25 13:30:00.023 MSFT 51.95 75</span>
<span class="go">1 2016-05-25 13:30:00.038 MSFT 51.95 155</span>
<span class="go">2 2016-05-25 13:30:00.048 GOOG 720.77 100</span>
<span class="go">3 2016-05-25 13:30:00.048 GOOG 720.92 100</span>
<span class="go">4 2016-05-25 13:30:00.048 AAPL 98.00 100</span>
<span class="gp">In [95]: </span><span class="n">quotes</span>
<span class="gr">Out[95]: </span>
<span class="go"> time ticker bid ask</span>
<span class="go">0 2016-05-25 13:30:00.023 GOOG 720.50 720.93</span>
<span class="go">1 2016-05-25 13:30:00.023 MSFT 51.95 51.96</span>
<span class="go">2 2016-05-25 13:30:00.030 MSFT 51.97 51.98</span>
<span class="go">3 2016-05-25 13:30:00.041 MSFT 51.99 52.00</span>
<span class="go">4 2016-05-25 13:30:00.048 GOOG 720.50 720.93</span>
<span class="go">5 2016-05-25 13:30:00.049 AAPL 97.99 98.01</span>
<span class="go">6 2016-05-25 13:30:00.072 GOOG 720.50 720.88</span>
<span class="go">7 2016-05-25 13:30:00.075 MSFT 52.01 52.03</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-283">默认情况下，我们使用asof的引号。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [96]: </span><span class="n">pd</span><span class="o">.</span><span class="n">merge_asof</span><span class="p">(</span><span class="n">trades</span><span class="p">,</span> <span class="n">quotes</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">on</span><span class="o">=</span><span class="s1">'time'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">by</span><span class="o">=</span><span class="s1">'ticker'</span><span class="p">)</span>
<span class="gp"> ....:</span>
<span class="gr">Out[96]: </span>
<span class="go"> time ticker price quantity bid ask</span>
<span class="go">0 2016-05-25 13:30:00.023 MSFT 51.95 75 51.95 51.96</span>
<span class="go">1 2016-05-25 13:30:00.038 MSFT 51.95 155 51.97 51.98</span>
<span class="go">2 2016-05-25 13:30:00.048 GOOG 720.77 100 720.50 720.93</span>
<span class="go">3 2016-05-25 13:30:00.048 GOOG 720.92 100 720.50 720.93</span>
<span class="go">4 2016-05-25 13:30:00.048 AAPL 98.00 100 NaN NaN</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-284">我们只在<code class="docutils literal"><span class="pre">2ms</span></code>之内的报价时间和交易时间之间。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [97]: </span><span class="n">pd</span><span class="o">.</span><span class="n">merge_asof</span><span class="p">(</span><span class="n">trades</span><span class="p">,</span> <span class="n">quotes</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">on</span><span class="o">=</span><span class="s1">'time'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">by</span><span class="o">=</span><span class="s1">'ticker'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">tolerance</span><span class="o">=</span><span class="n">pd</span><span class="o">.</span><span class="n">Timedelta</span><span class="p">(</span><span class="s1">'2ms'</span><span class="p">))</span>
<span class="gp"> ....:</span>
<span class="gr">Out[97]: </span>
<span class="go"> time ticker price quantity bid ask</span>
<span class="go">0 2016-05-25 13:30:00.023 MSFT 51.95 75 51.95 51.96</span>
<span class="go">1 2016-05-25 13:30:00.038 MSFT 51.95 155 NaN NaN</span>
<span class="go">2 2016-05-25 13:30:00.048 GOOG 720.77 100 720.50 720.93</span>
<span class="go">3 2016-05-25 13:30:00.048 GOOG 720.92 100 720.50 720.93</span>
<span class="go">4 2016-05-25 13:30:00.048 AAPL 98.00 100 NaN NaN</span>
</pre></div>
</div>
<p><span class="yiyi-st" id="yiyi-285">我们只有在<code class="docutils literal"><span class="pre">10ms</span></code>之内的报价时间和交易时间之间，我们排除准时匹配。</span><span class="yiyi-st" id="yiyi-286">注意，虽然我们排除了（报价）的精确匹配，但是先前报价传播到该时间点。</span></p>
<div class="highlight-ipython"><div class="highlight"><pre><span></span><span class="gp">In [98]: </span><span class="n">pd</span><span class="o">.</span><span class="n">merge_asof</span><span class="p">(</span><span class="n">trades</span><span class="p">,</span> <span class="n">quotes</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">on</span><span class="o">=</span><span class="s1">'time'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">by</span><span class="o">=</span><span class="s1">'ticker'</span><span class="p">,</span>
<span class="gp"> ....:</span> <span class="n">tolerance</span><span class="o">=</span><span class="n">pd</span><span class="o">.</span><span class="n">Timedelta</span><span class="p">(</span><span class="s1">'10ms'</span><span class="p">),</span>
<span class="gp"> ....:</span> <span class="n">allow_exact_matches</span><span class="o">=</span><span class="bp">False</span><span class="p">)</span>
<span class="gp"> ....:</span>
<span class="gr">Out[98]: </span>
<span class="go"> time ticker price quantity bid ask</span>
<span class="go">0 2016-05-25 13:30:00.023 MSFT 51.95 75 NaN NaN</span>
<span class="go">1 2016-05-25 13:30:00.038 MSFT 51.95 155 51.97 51.98</span>
<span class="go">2 2016-05-25 13:30:00.048 GOOG 720.77 100 NaN NaN</span>
<span class="go">3 2016-05-25 13:30:00.048 GOOG 720.92 100 NaN NaN</span>
<span class="go">4 2016-05-25 13:30:00.048 AAPL 98.00 100 NaN NaN</span>
</pre></div>
</div>
</div>
</div>