This repository has been archived by the owner on May 6, 2021. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 87
/
10min.html
1015 lines (943 loc) · 164 KB
/
10min.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<h1>十分钟搞定 pandas</h1>
<blockquote>
<p>原文：<a href="http://pandas.pydata.org/pandas-docs/stable/10min.html">http://pandas.pydata.org/pandas-docs/stable/10min.html</a></p>
<p>译者：<a href="http://home.cnblogs.com/u/chaosimple/">ChaoSimple</a></p>
<p>校对：<a href="https://github.com/wizardforcel">飞龙</a></p>
</blockquote>
<p>官方网站上《10 Minutes to pandas》的一个简单的翻译，原文在<a href="http://pandas.pydata.org/pandas-docs/stable/10min.html" rel="nofollow">这里</a>。这篇文章是对 pandas 的一个简单的介绍，详细的介绍请参考：<a href="http://pandas.pydata.org/pandas-docs/stable/cookbook.html#cookbook" rel="nofollow">秘籍</a> 。习惯上，我们会按下面格式引入所需要的包：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">1</span>]: <span class="pl-k">import</span> pandas <span class="pl-k">as</span> pd
In [<span class="pl-c1">2</span>]: <span class="pl-k">import</span> numpy <span class="pl-k">as</span> np
In [<span class="pl-c1">3</span>]: <span class="pl-k">import</span> matplotlib.pyplot <span class="pl-k">as</span> plt</pre></div>
<h1>一、 创建对象</h1>
<p>可以通过 <a href="http://pandas.pydata.org/pandas-docs/stable/dsintro.html#dsintro" rel="nofollow">数据结构入门</a> 来查看有关该节内容的详细信息。</p>
<p>1、可以通过传递一个<code>list</code>对象来创建一个<code>Series</code>，pandas 会默认创建整型索引：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">4</span>]: s <span class="pl-k">=</span> pd.Series([<span class="pl-c1">1</span>,<span class="pl-c1">3</span>,<span class="pl-c1">5</span>,np.nan,<span class="pl-c1">6</span>,<span class="pl-c1">8</span>])
In [<span class="pl-c1">5</span>]: s
Out[<span class="pl-c1">5</span>]:
<span class="pl-c1">0</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">1</span> <span class="pl-c1">3.0</span>
<span class="pl-c1">2</span> <span class="pl-c1">5.0</span>
<span class="pl-c1">3</span> NaN
<span class="pl-c1">4</span> <span class="pl-c1">6.0</span>
<span class="pl-c1">5</span> <span class="pl-c1">8.0</span>
dtype: float64</pre></div>
<p>2、通过传递一个 numpy<code>array</code>，时间索引以及列标签来创建一个<code>DataFrame</code>：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">6</span>]: dates <span class="pl-k">=</span> pd.date_range(<span class="pl-s"><span class="pl-pds">'</span>20130101<span class="pl-pds">'</span></span>, <span class="pl-v">periods</span><span class="pl-k">=</span><span class="pl-c1">6</span>)
In [<span class="pl-c1">7</span>]: dates
Out[<span class="pl-c1">7</span>]:
DatetimeIndex([<span class="pl-s"><span class="pl-pds">'</span>2013-01-01<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-02<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-03<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-04<span class="pl-pds">'</span></span>,
<span class="pl-s"><span class="pl-pds">'</span>2013-01-05<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-06<span class="pl-pds">'</span></span>],
<span class="pl-v">dtype</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>datetime64[ns]<span class="pl-pds">'</span></span>, <span class="pl-v">freq</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">8</span>]: df <span class="pl-k">=</span> pd.DataFrame(np.random.randn(<span class="pl-c1">6</span>,<span class="pl-c1">4</span>), <span class="pl-v">index</span><span class="pl-k">=</span>dates, <span class="pl-v">columns</span><span class="pl-k">=</span><span class="pl-c1">list</span>(<span class="pl-s"><span class="pl-pds">'</span>ABCD<span class="pl-pds">'</span></span>))
In [<span class="pl-c1">9</span>]: df
Out[<span class="pl-c1">9</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-c1">0.524988</span></pre></div>
<p>3、通过传递一个能够被转换成类似序列结构的字典对象来创建一个<code>DataFrame</code>：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">10</span>]: df2 <span class="pl-k">=</span> pd.DataFrame({ <span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span> : <span class="pl-c1">1</span>.,
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span> : pd.Timestamp(<span class="pl-s"><span class="pl-pds">'</span>20130102<span class="pl-pds">'</span></span>),
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span> : pd.Series(<span class="pl-c1">1</span>,<span class="pl-v">index</span><span class="pl-k">=</span><span class="pl-c1">list</span>(<span class="pl-c1">range</span>(<span class="pl-c1">4</span>)),<span class="pl-v">dtype</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>float32<span class="pl-pds">'</span></span>),
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span> : np.array([<span class="pl-c1">3</span>] <span class="pl-k">*</span> <span class="pl-c1">4</span>,<span class="pl-v">dtype</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>int32<span class="pl-pds">'</span></span>),
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>E<span class="pl-pds">'</span></span> : pd.Categorical([<span class="pl-s"><span class="pl-pds">"</span>test<span class="pl-pds">"</span></span>,<span class="pl-s"><span class="pl-pds">"</span>train<span class="pl-pds">"</span></span>,<span class="pl-s"><span class="pl-pds">"</span>test<span class="pl-pds">"</span></span>,<span class="pl-s"><span class="pl-pds">"</span>train<span class="pl-pds">"</span></span>]),
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>F<span class="pl-pds">'</span></span> : <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span> })
<span class="pl-c1">...</span>.:
In [<span class="pl-c1">11</span>]: df2
Out[<span class="pl-c1">11</span>]:
A B C D E F
<span class="pl-c1">0</span> <span class="pl-c1">1.0</span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.0</span> <span class="pl-c1">3</span> test foo
<span class="pl-c1">1</span> <span class="pl-c1">1.0</span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.0</span> <span class="pl-c1">3</span> train foo
<span class="pl-c1">2</span> <span class="pl-c1">1.0</span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.0</span> <span class="pl-c1">3</span> test foo
<span class="pl-c1">3</span> <span class="pl-c1">1.0</span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.0</span> <span class="pl-c1">3</span> train foo</pre></div>
<p>4、查看不同列的数据类型：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">12</span>]: df2.dtypes
Out[<span class="pl-c1">12</span>]:
A float64
B datetime64[ns]
C float32
D int32
E category
F <span class="pl-c1">object</span>
dtype: <span class="pl-c1">object</span></pre></div>
<p>5、如果你使用的是 IPython，使用 Tab 自动补全功能会自动识别所有的属性以及自定义的列，下图中是所有能够被自动识别的属性的一个子集：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">13</span>]: df2.<span class="pl-k"><</span><span class="pl-c1">TAB</span><span class="pl-k">></span>
df2.A df2.boxplot
df2.abs df2.C
df2.add df2.clip
df2.add_prefix df2.clip_lower
df2.add_suffix df2.clip_upper
df2.align df2.columns
df2.all df2.combine
df2.any df2.combineAdd
df2.append df2.combine_first
df2.apply df2.combineMult
df2.applymap df2.compound
df2.as_blocks df2.consolidate
df2.asfreq df2.convert_objects
df2.as_matrix df2.copy
df2.astype df2.corr
df2.at df2.corrwith
df2.at_time df2.count
df2.axes df2.cov
df2.B df2.cummax
df2.between_time df2.cummin
df2.bfill df2.cumprod
df2.blocks df2.cumsum
df2.bool df2.D</pre></div>
<h1>二、 查看数据</h1>
<p>详情请参阅：<a href="http://pandas.pydata.org/pandas-docs/stable/basics.html#basics" rel="nofollow">基础</a>。</p>
<p>1、 查看<code>DataFrame</code>中头部和尾部的行：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">14</span>]: df.head()
Out[<span class="pl-c1">14</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span>
In [<span class="pl-c1">15</span>]: df.tail(<span class="pl-c1">3</span>)
Out[<span class="pl-c1">15</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-c1">0.524988</span></pre></div>
<p>2、 显示索引、列和底层的 numpy 数据：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">16</span>]: df.index
Out[<span class="pl-c1">16</span>]:
DatetimeIndex([<span class="pl-s"><span class="pl-pds">'</span>2013-01-01<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-02<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-03<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-04<span class="pl-pds">'</span></span>,
<span class="pl-s"><span class="pl-pds">'</span>2013-01-05<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2013-01-06<span class="pl-pds">'</span></span>],
<span class="pl-v">dtype</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>datetime64[ns]<span class="pl-pds">'</span></span>, <span class="pl-v">freq</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">17</span>]: df.columns
Out[<span class="pl-c1">17</span>]: Index([<span class="pl-s"><span class="pl-k">u</span><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-k">u</span><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-k">u</span><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-k">u</span><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>], <span class="pl-v">dtype</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>object<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">18</span>]: df.values
Out[<span class="pl-c1">18</span>]:
array([[ <span class="pl-c1">0.4691</span>, <span class="pl-k">-</span><span class="pl-c1">0.2829</span>, <span class="pl-k">-</span><span class="pl-c1">1.5091</span>, <span class="pl-k">-</span><span class="pl-c1">1.1356</span>],
[ <span class="pl-c1">1.2121</span>, <span class="pl-k">-</span><span class="pl-c1">0.1732</span>, <span class="pl-c1">0.1192</span>, <span class="pl-k">-</span><span class="pl-c1">1.0442</span>],
[<span class="pl-k">-</span><span class="pl-c1">0.8618</span>, <span class="pl-k">-</span><span class="pl-c1">2.1046</span>, <span class="pl-k">-</span><span class="pl-c1">0.4949</span>, <span class="pl-c1">1.0718</span>],
[ <span class="pl-c1">0.7216</span>, <span class="pl-k">-</span><span class="pl-c1">0.7068</span>, <span class="pl-k">-</span><span class="pl-c1">1.0396</span>, <span class="pl-c1">0.2719</span>],
[<span class="pl-k">-</span><span class="pl-c1">0.425</span> , <span class="pl-c1">0.567</span> , <span class="pl-c1">0.2762</span>, <span class="pl-k">-</span><span class="pl-c1">1.0874</span>],
[<span class="pl-k">-</span><span class="pl-c1">0.6737</span>, <span class="pl-c1">0.1136</span>, <span class="pl-k">-</span><span class="pl-c1">1.4784</span>, <span class="pl-c1">0.525</span> ]])</pre></div>
<p>3、 <code>describe()</code>函数对于数据的快速统计汇总：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">19</span>]: df.describe()
Out[<span class="pl-c1">19</span>]:
A B C D
count <span class="pl-c1">6.000000</span> <span class="pl-c1">6.000000</span> <span class="pl-c1">6.000000</span> <span class="pl-c1">6.000000</span>
mean <span class="pl-c1">0.073711</span> <span class="pl-k">-</span><span class="pl-c1">0.431125</span> <span class="pl-k">-</span><span class="pl-c1">0.687758</span> <span class="pl-k">-</span><span class="pl-c1">0.233103</span>
std <span class="pl-c1">0.843157</span> <span class="pl-c1">0.922818</span> <span class="pl-c1">0.779887</span> <span class="pl-c1">0.973118</span>
<span class="pl-c1">min</span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
<span class="pl-c1">25</span><span class="pl-k">%</span> <span class="pl-k">-</span><span class="pl-c1">0.611510</span> <span class="pl-k">-</span><span class="pl-c1">0.600794</span> <span class="pl-k">-</span><span class="pl-c1">1.368714</span> <span class="pl-k">-</span><span class="pl-c1">1.076610</span>
<span class="pl-c1">50</span><span class="pl-k">%</span> <span class="pl-c1">0.022070</span> <span class="pl-k">-</span><span class="pl-c1">0.228039</span> <span class="pl-k">-</span><span class="pl-c1">0.767252</span> <span class="pl-k">-</span><span class="pl-c1">0.386188</span>
<span class="pl-c1">75</span><span class="pl-k">%</span> <span class="pl-c1">0.658444</span> <span class="pl-c1">0.041933</span> <span class="pl-k">-</span><span class="pl-c1">0.034326</span> <span class="pl-c1">0.461706</span>
<span class="pl-c1">max</span> <span class="pl-c1">1.212112</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-c1">1.071804</span></pre></div>
<p>4、 对数据的转置：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">20</span>]: df.T
Out[<span class="pl-c1">20</span>]:
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span>
A <span class="pl-c1">0.469112</span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span>
B <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.113648</span>
C <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span>
D <span class="pl-k">-</span><span class="pl-c1">1.135632</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span> <span class="pl-c1">1.071804</span> <span class="pl-c1">0.271860</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span> <span class="pl-c1">0.524988</span></pre></div>
<p>5、 按轴进行排序</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">21</span>]: df.sort_index(<span class="pl-v">axis</span><span class="pl-k">=</span><span class="pl-c1">1</span>, <span class="pl-v">ascending</span><span class="pl-k">=</span><span class="pl-c1">False</span>)
Out[<span class="pl-c1">21</span>]:
D C B A
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-c1">0.469112</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">1.212112</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">1.071804</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.271860</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-c1">0.721555</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span> <span class="pl-c1">0.276232</span> <span class="pl-c1">0.567020</span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">0.524988</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span></pre></div>
<p>6、 按值进行排序</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">22</span>]: df.sort_values(<span class="pl-v">by</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">22</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-c1">0.524988</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span></pre></div>
<h1>三、 选择</h1>
<p>虽然标准的 Python/Numpy 的选择和设置表达式都能够直接派上用场，但是作为工程使用的代码，我们推荐使用经过优化的 pandas 数据访问方式： <code>.at</code>, <code>.iat</code>, <code>.loc</code>, <code>.iloc</code> 和 <code>.ix</code>。详情请参阅<a href="http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing" rel="nofollow">索引和选取数据</a> 和 <a href="http://pandas.pydata.org/pandas-docs/stable/advanced.html#advanced" rel="nofollow">多重索引/高级索引</a>。</p>
<h2>获取</h2>
<p>1、 选择一个单独的列，这将会返回一个<code>Series</code>，等同于<code>df.A</code>：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">23</span>]: df[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>]
Out[<span class="pl-c1">23</span>]:
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span>
Freq: D, Name: A, dtype: float64</pre></div>
<p>2、 通过<code>[]</code>进行选择，这将会对行进行切片</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">24</span>]: df[<span class="pl-c1">0</span>:<span class="pl-c1">3</span>]
Out[<span class="pl-c1">24</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span>
In [<span class="pl-c1">25</span>]: df[<span class="pl-s"><span class="pl-pds">'</span>20130102<span class="pl-pds">'</span></span>:<span class="pl-s"><span class="pl-pds">'</span>20130104<span class="pl-pds">'</span></span>]
Out[<span class="pl-c1">25</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span></pre></div>
<h2>通过标签选择</h2>
<p>1、 使用标签来获取一个交叉的区域</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">26</span>]: df.loc[dates[<span class="pl-c1">0</span>]]
Out[<span class="pl-c1">26</span>]:
A <span class="pl-c1">0.469112</span>
B <span class="pl-k">-</span><span class="pl-c1">0.282863</span>
C <span class="pl-k">-</span><span class="pl-c1">1.509059</span>
D <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
Name: <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span>, dtype: float64</pre></div>
<p>2、 通过标签来在多个轴上进行选择</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">27</span>]: df.loc[:,[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>]]
Out[<span class="pl-c1">27</span>]:
A B
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-c1">0.113648</span></pre></div>
<p>3、 标签切片</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">28</span>]: df.loc[<span class="pl-s"><span class="pl-pds">'</span>20130102<span class="pl-pds">'</span></span>:<span class="pl-s"><span class="pl-pds">'</span>20130104<span class="pl-pds">'</span></span>,[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>]]
Out[<span class="pl-c1">28</span>]:
A B
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span></pre></div>
<p>4、 对于返回的对象进行维度缩减</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">29</span>]: df.loc[<span class="pl-s"><span class="pl-pds">'</span>20130102<span class="pl-pds">'</span></span>,[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>]]
Out[<span class="pl-c1">29</span>]:
A <span class="pl-c1">1.212112</span>
B <span class="pl-k">-</span><span class="pl-c1">0.173215</span>
Name: <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span>, dtype: float64</pre></div>
<p>5、 获取一个标量</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">30</span>]: df.loc[dates[<span class="pl-c1">0</span>],<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>]
Out[<span class="pl-c1">30</span>]: <span class="pl-c1">0.46911229990718628</span></pre></div>
<p>6、 快速访问一个标量（与上一个方法等价）</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">31</span>]: df.at[dates[<span class="pl-c1">0</span>],<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>]
Out[<span class="pl-c1">31</span>]: <span class="pl-c1">0.46911229990718628</span></pre></div>
<h2>通过位置选择</h2>
<p>1、 通过传递数值进行位置选择（选择的是行）</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">32</span>]: df.iloc[<span class="pl-c1">3</span>]
Out[<span class="pl-c1">32</span>]:
A <span class="pl-c1">0.721555</span>
B <span class="pl-k">-</span><span class="pl-c1">0.706771</span>
C <span class="pl-k">-</span><span class="pl-c1">1.039575</span>
D <span class="pl-c1">0.271860</span>
Name: <span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span>, dtype: float64</pre></div>
<p>2、 通过数值进行切片，与 numpy/python 中的情况类似</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">33</span>]: df.iloc[<span class="pl-c1">3</span>:<span class="pl-c1">5</span>,<span class="pl-c1">0</span>:<span class="pl-c1">2</span>]
Out[<span class="pl-c1">33</span>]:
A B
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span></pre></div>
<p>3、 通过指定一个位置的列表，与 numpy/python 中的情况类似</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">34</span>]: df.iloc[[<span class="pl-c1">1</span>,<span class="pl-c1">2</span>,<span class="pl-c1">4</span>],[<span class="pl-c1">0</span>,<span class="pl-c1">2</span>]]
Out[<span class="pl-c1">34</span>]:
A C
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-c1">0.119209</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.276232</span></pre></div>
<p>4、 对行进行切片</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">35</span>]: df.iloc[<span class="pl-c1">1</span>:<span class="pl-c1">3</span>,:]
Out[<span class="pl-c1">35</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span></pre></div>
<p>5、 对列进行切片</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">36</span>]: df.iloc[:,<span class="pl-c1">1</span>:<span class="pl-c1">3</span>]
Out[<span class="pl-c1">36</span>]:
B C
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span></pre></div>
<p>6、 获取特定的值</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">37</span>]: df.iloc[<span class="pl-c1">1</span>,<span class="pl-c1">1</span>]
Out[<span class="pl-c1">37</span>]: <span class="pl-k">-</span><span class="pl-c1">0.17321464905330858</span></pre></div>
<p>快速访问标量（等同于前一个方法）：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">38</span>]: df.iat[<span class="pl-c1">1</span>,<span class="pl-c1">1</span>]
Out[<span class="pl-c1">38</span>]: <span class="pl-k">-</span><span class="pl-c1">0.17321464905330858</span></pre></div>
<h2>布尔索引</h2>
<p>1、 使用一个单独列的值来选择数据：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">39</span>]: df[df.A <span class="pl-k">></span> <span class="pl-c1">0</span>]
Out[<span class="pl-c1">39</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span></pre></div>
<p>2、 使用<code>where</code>操作来选择数据：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">40</span>]: df[df <span class="pl-k">></span> <span class="pl-c1">0</span>]
Out[<span class="pl-c1">40</span>]:
A B C D
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> NaN NaN NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> NaN <span class="pl-c1">0.119209</span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> NaN NaN NaN <span class="pl-c1">1.071804</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> NaN NaN <span class="pl-c1">0.271860</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> NaN <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> NaN <span class="pl-c1">0.113648</span> NaN <span class="pl-c1">0.524988</span></pre></div>
<p>3、 使用<code>isin()</code>方法来过滤：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">41</span>]: df2 <span class="pl-k">=</span> df.copy()
In [<span class="pl-c1">42</span>]: df2[<span class="pl-s"><span class="pl-pds">'</span>E<span class="pl-pds">'</span></span>] <span class="pl-k">=</span> [<span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>three<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>four<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>three<span class="pl-pds">'</span></span>]
In [<span class="pl-c1">43</span>]: df2
Out[<span class="pl-c1">43</span>]:
A B C D E
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.469112</span> <span class="pl-k">-</span><span class="pl-c1">0.282863</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">1.135632</span> one
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">1.044236</span> one
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span> two
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">0.271860</span> three
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span> four
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-c1">0.524988</span> three
In [<span class="pl-c1">44</span>]: df2[df2[<span class="pl-s"><span class="pl-pds">'</span>E<span class="pl-pds">'</span></span>].isin([<span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>four<span class="pl-pds">'</span></span>])]
Out[<span class="pl-c1">44</span>]:
A B C D E
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">1.071804</span> two
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">1.087401</span> four</pre></div>
<h2>设置</h2>
<p>1、 设置一个新的列：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">45</span>]: s1 <span class="pl-k">=</span> pd.Series([<span class="pl-c1">1</span>,<span class="pl-c1">2</span>,<span class="pl-c1">3</span>,<span class="pl-c1">4</span>,<span class="pl-c1">5</span>,<span class="pl-c1">6</span>], <span class="pl-v">index</span><span class="pl-k">=</span>pd.date_range(<span class="pl-s"><span class="pl-pds">'</span>20130102<span class="pl-pds">'</span></span>, <span class="pl-v">periods</span><span class="pl-k">=</span><span class="pl-c1">6</span>))
In [<span class="pl-c1">46</span>]: s1
Out[<span class="pl-c1">46</span>]:
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">2</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">3</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">4</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">5</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">6</span>
Freq: D, dtype: int64
In [<span class="pl-c1">47</span>]: df[<span class="pl-s"><span class="pl-pds">'</span>F<span class="pl-pds">'</span></span>] <span class="pl-k">=</span> s1</pre></div>
<p>2、 通过标签设置新的值：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">48</span>]: df.at[dates[<span class="pl-c1">0</span>],<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>] <span class="pl-k">=</span> <span class="pl-c1">0</span></pre></div>
<p>3、 通过位置设置新的值：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">49</span>]: df.iat[<span class="pl-c1">0</span>,<span class="pl-c1">1</span>] <span class="pl-k">=</span> <span class="pl-c1">0</span></pre></div>
<p>4、 通过一个numpy数组设置一组新值：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">50</span>]: df.loc[:,<span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>] <span class="pl-k">=</span> np.array([<span class="pl-c1">5</span>] <span class="pl-k">*</span> <span class="pl-c1">len</span>(df))</pre></div>
<p>上述操作结果如下：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">51</span>]: df
Out[<span class="pl-c1">51</span>]:
A B C D F
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.000000</span> <span class="pl-c1">0.000000</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-c1">5</span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-c1">5</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">5</span> <span class="pl-c1">2.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">5</span> <span class="pl-c1">3.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-c1">0.567020</span> <span class="pl-c1">0.276232</span> <span class="pl-c1">5</span> <span class="pl-c1">4.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-c1">5</span> <span class="pl-c1">5.0</span></pre></div>
<p>5、 通过where操作来设置新的值：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">52</span>]: df2 <span class="pl-k">=</span> df.copy()
In [<span class="pl-c1">53</span>]: df2[df2 <span class="pl-k">></span> <span class="pl-c1">0</span>] <span class="pl-k">=</span> <span class="pl-k">-</span>df2
In [<span class="pl-c1">54</span>]: df2
Out[<span class="pl-c1">54</span>]:
A B C D F
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.000000</span> <span class="pl-c1">0.000000</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-k">-</span><span class="pl-c1">5</span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-k">-</span><span class="pl-c1">0.119209</span> <span class="pl-k">-</span><span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-k">-</span><span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">2.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-k">-</span><span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-k">-</span><span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">3.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">0.424972</span> <span class="pl-k">-</span><span class="pl-c1">0.567020</span> <span class="pl-k">-</span><span class="pl-c1">0.276232</span> <span class="pl-k">-</span><span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">4.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.673690</span> <span class="pl-k">-</span><span class="pl-c1">0.113648</span> <span class="pl-k">-</span><span class="pl-c1">1.478427</span> <span class="pl-k">-</span><span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">5.0</span></pre></div>
<h1>四、 缺失值处理</h1>
<p>在 pandas 中，使用<code>np.nan</code>来代替缺失值，这些值将默认不会包含在计算中，详情请参阅：<a href="http://pandas.pydata.org/pandas-docs/stable/missing_data.html#missing-data" rel="nofollow">缺失的数据</a>。</p>
<p>1、 <code>reindex()</code>方法可以对指定轴上的索引进行改变/增加/删除操作，这将返回原始数据的一个拷贝：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">55</span>]: df1 <span class="pl-k">=</span> df.reindex(<span class="pl-v">index</span><span class="pl-k">=</span>dates[<span class="pl-c1">0</span>:<span class="pl-c1">4</span>], <span class="pl-v">columns</span><span class="pl-k">=</span><span class="pl-c1">list</span>(df.columns) <span class="pl-k">+</span> [<span class="pl-s"><span class="pl-pds">'</span>E<span class="pl-pds">'</span></span>])
In [<span class="pl-c1">56</span>]: df1.loc[dates[<span class="pl-c1">0</span>]:dates[<span class="pl-c1">1</span>],<span class="pl-s"><span class="pl-pds">'</span>E<span class="pl-pds">'</span></span>] <span class="pl-k">=</span> <span class="pl-c1">1</span>
In [<span class="pl-c1">57</span>]: df1
Out[<span class="pl-c1">57</span>]:
A B C D F E
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.000000</span> <span class="pl-c1">0.000000</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-c1">5</span> NaN <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-c1">5</span> <span class="pl-c1">1.0</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">5</span> <span class="pl-c1">2.0</span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">5</span> <span class="pl-c1">3.0</span> NaN</pre></div>
<p>2、 去掉包含缺失值的行：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">58</span>]: df1.dropna(<span class="pl-v">how</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>any<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">58</span>]:
A B C D F E
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-c1">5</span> <span class="pl-c1">1.0</span> <span class="pl-c1">1.0</span></pre></div>
<p>3、 对缺失值进行填充：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">59</span>]: df1.fillna(<span class="pl-v">value</span><span class="pl-k">=</span><span class="pl-c1">5</span>)
Out[<span class="pl-c1">59</span>]:
A B C D F E
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.000000</span> <span class="pl-c1">0.000000</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-c1">5</span> <span class="pl-c1">5.0</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-c1">0.119209</span> <span class="pl-c1">5</span> <span class="pl-c1">1.0</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.861849</span> <span class="pl-k">-</span><span class="pl-c1">2.104569</span> <span class="pl-k">-</span><span class="pl-c1">0.494929</span> <span class="pl-c1">5</span> <span class="pl-c1">2.0</span> <span class="pl-c1">5.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.721555</span> <span class="pl-k">-</span><span class="pl-c1">0.706771</span> <span class="pl-k">-</span><span class="pl-c1">1.039575</span> <span class="pl-c1">5</span> <span class="pl-c1">3.0</span> <span class="pl-c1">5.0</span></pre></div>
<p>4、 对数据进行布尔填充：</p>
<div class="highlight highlight-source-python"><pre>n [<span class="pl-c1">60</span>]: pd.isnull(df1)
Out[<span class="pl-c1">60</span>]:
A B C D F E
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">True</span> <span class="pl-c1">False</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">True</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">False</span> <span class="pl-c1">True</span></pre></div>
<h1>五、 相关操作</h1>
<p>详情请参与 <a href="http://pandas.pydata.org/pandas-docs/stable/basics.html#basics-binop" rel="nofollow">基本的二进制操作</a></p>
<h2>统计（相关操作通常情况下不包括缺失值）</h2>
<p>1、 执行描述性统计：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">61</span>]: df.mean()
Out[<span class="pl-c1">61</span>]:
A <span class="pl-k">-</span><span class="pl-c1">0.004474</span>
B <span class="pl-k">-</span><span class="pl-c1">0.383981</span>
C <span class="pl-k">-</span><span class="pl-c1">0.687758</span>
D <span class="pl-c1">5.000000</span>
F <span class="pl-c1">3.000000</span>
dtype: float64</pre></div>
<p>2、 在其他轴上进行相同的操作：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">62</span>]: df.mean(<span class="pl-c1">1</span>)
Out[<span class="pl-c1">62</span>]:
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.872735</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.431621</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">0.707731</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">1.395042</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">1.883656</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">1.592306</span>
Freq: D, dtype: float64</pre></div>
<p>3、 对于拥有不同维度，需要对齐的对象进行操作。Pandas 会自动的沿着指定的维度进行广播：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">63</span>]: s <span class="pl-k">=</span> pd.Series([<span class="pl-c1">1</span>,<span class="pl-c1">3</span>,<span class="pl-c1">5</span>,np.nan,<span class="pl-c1">6</span>,<span class="pl-c1">8</span>], <span class="pl-v">index</span><span class="pl-k">=</span>dates).shift(<span class="pl-c1">2</span>)
In [<span class="pl-c1">64</span>]: s
Out[<span class="pl-c1">64</span>]:
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">3.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">5.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> NaN
Freq: D, dtype: float64
In [<span class="pl-c1">65</span>]: df.sub(s, <span class="pl-v">axis</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>index<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">65</span>]:
A B C D F
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> NaN NaN NaN NaN NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> NaN NaN NaN NaN NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">1.861849</span> <span class="pl-k">-</span><span class="pl-c1">3.104569</span> <span class="pl-k">-</span><span class="pl-c1">1.494929</span> <span class="pl-c1">4.0</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-k">-</span><span class="pl-c1">2.278445</span> <span class="pl-k">-</span><span class="pl-c1">3.706771</span> <span class="pl-k">-</span><span class="pl-c1">4.039575</span> <span class="pl-c1">2.0</span> <span class="pl-c1">0.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-k">-</span><span class="pl-c1">5.424972</span> <span class="pl-k">-</span><span class="pl-c1">4.432980</span> <span class="pl-k">-</span><span class="pl-c1">4.723768</span> <span class="pl-c1">0.0</span> <span class="pl-k">-</span><span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> NaN NaN NaN NaN NaN</pre></div>
<h2><code>Apply</code></h2>
<p>1、 对数据应用函数：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">66</span>]: df.apply(np.cumsum)
Out[<span class="pl-c1">66</span>]:
A B C D F
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.000000</span> <span class="pl-c1">0.000000</span> <span class="pl-k">-</span><span class="pl-c1">1.509059</span> <span class="pl-c1">5</span> NaN
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-c1">1.212112</span> <span class="pl-k">-</span><span class="pl-c1">0.173215</span> <span class="pl-k">-</span><span class="pl-c1">1.389850</span> <span class="pl-c1">10</span> <span class="pl-c1">1.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-c1">0.350263</span> <span class="pl-k">-</span><span class="pl-c1">2.277784</span> <span class="pl-k">-</span><span class="pl-c1">1.884779</span> <span class="pl-c1">15</span> <span class="pl-c1">3.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">1.071818</span> <span class="pl-k">-</span><span class="pl-c1">2.984555</span> <span class="pl-k">-</span><span class="pl-c1">2.924354</span> <span class="pl-c1">20</span> <span class="pl-c1">6.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">0.646846</span> <span class="pl-k">-</span><span class="pl-c1">2.417535</span> <span class="pl-k">-</span><span class="pl-c1">2.648122</span> <span class="pl-c1">25</span> <span class="pl-c1">10.0</span>
<span class="pl-c1">2013</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-k">-</span><span class="pl-c1">0.026844</span> <span class="pl-k">-</span><span class="pl-c1">2.303886</span> <span class="pl-k">-</span><span class="pl-c1">4.126549</span> <span class="pl-c1">30</span> <span class="pl-c1">15.0</span>
In [<span class="pl-c1">67</span>]: df.apply(<span class="pl-k">lambda</span> <span class="pl-smi">x</span>: x.max() <span class="pl-k">-</span> x.min())
Out[<span class="pl-c1">67</span>]:
A <span class="pl-c1">2.073961</span>
B <span class="pl-c1">2.671590</span>
C <span class="pl-c1">1.785291</span>
D <span class="pl-c1">0.000000</span>
F <span class="pl-c1">4.000000</span>
dtype: float64</pre></div>
<h2>直方图</h2>
<p>具体请参照：<a href="http://pandas.pydata.org/pandas-docs/stable/basics.html#basics-discretization" rel="nofollow">直方图和离散化</a>。</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">68</span>]: s <span class="pl-k">=</span> pd.Series(np.random.randint(<span class="pl-c1">0</span>, <span class="pl-c1">7</span>, <span class="pl-v">size</span><span class="pl-k">=</span><span class="pl-c1">10</span>))
In [<span class="pl-c1">69</span>]: s
Out[<span class="pl-c1">69</span>]:
<span class="pl-c1">0</span> <span class="pl-c1">4</span>
<span class="pl-c1">1</span> <span class="pl-c1">2</span>
<span class="pl-c1">2</span> <span class="pl-c1">1</span>
<span class="pl-c1">3</span> <span class="pl-c1">2</span>
<span class="pl-c1">4</span> <span class="pl-c1">6</span>
<span class="pl-c1">5</span> <span class="pl-c1">4</span>
<span class="pl-c1">6</span> <span class="pl-c1">4</span>
<span class="pl-c1">7</span> <span class="pl-c1">6</span>
<span class="pl-c1">8</span> <span class="pl-c1">4</span>
<span class="pl-c1">9</span> <span class="pl-c1">4</span>
dtype: int64
In [<span class="pl-c1">70</span>]: s.value_counts()
Out[<span class="pl-c1">70</span>]:
<span class="pl-c1">4</span> <span class="pl-c1">5</span>
<span class="pl-c1">6</span> <span class="pl-c1">2</span>
<span class="pl-c1">2</span> <span class="pl-c1">2</span>
<span class="pl-c1">1</span> <span class="pl-c1">1</span>
dtype: int64</pre></div>
<h2>字符串方法</h2>
<p><code>Series</code>对象在其<code>str</code>属性中配备了一组字符串处理方法，可以很容易的应用到数组中的每个元素，如下段代码所示。更多详情请参考：<a href="http://pandas.pydata.org/pandas-docs/stable/text.html#text-string-methods" rel="nofollow">字符串向量化方法</a>。</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">71</span>]: s <span class="pl-k">=</span> pd.Series([<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>Aaba<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>Baca<span class="pl-pds">'</span></span>, np.nan, <span class="pl-s"><span class="pl-pds">'</span>CABA<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>dog<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>cat<span class="pl-pds">'</span></span>])
In [<span class="pl-c1">72</span>]: s.str.lower()
Out[<span class="pl-c1">72</span>]:
<span class="pl-c1">0</span> a
<span class="pl-c1">1</span> b
<span class="pl-c1">2</span> c
<span class="pl-c1">3</span> aaba
<span class="pl-c1">4</span> baca
<span class="pl-c1">5</span> NaN
<span class="pl-c1">6</span> caba
<span class="pl-c1">7</span> dog
<span class="pl-c1">8</span> cat
dtype: <span class="pl-c1">object</span></pre></div>
<h1>六、 合并</h1>
<p>Pandas 提供了大量的方法能够轻松的对<code>Series</code>，<code>DataFrame</code>和<code>Panel</code>对象进行各种符合各种逻辑关系的合并操作。具体请参阅：<a href="http://pandas.pydata.org/pandas-docs/stable/merging.html#merging" rel="nofollow">合并</a>。</p>
<h2><code>Concat</code></h2>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">73</span>]: df <span class="pl-k">=</span> pd.DataFrame(np.random.randn(<span class="pl-c1">10</span>, <span class="pl-c1">4</span>))
In [<span class="pl-c1">74</span>]: df
Out[<span class="pl-c1">74</span>]:
<span class="pl-c1">0</span> <span class="pl-c1">1</span> <span class="pl-c1">2</span> <span class="pl-c1">3</span>
<span class="pl-c1">0</span> <span class="pl-k">-</span><span class="pl-c1">0.548702</span> <span class="pl-c1">1.467327</span> <span class="pl-k">-</span><span class="pl-c1">1.015962</span> <span class="pl-k">-</span><span class="pl-c1">0.483075</span>
<span class="pl-c1">1</span> <span class="pl-c1">1.637550</span> <span class="pl-k">-</span><span class="pl-c1">1.217659</span> <span class="pl-k">-</span><span class="pl-c1">0.291519</span> <span class="pl-k">-</span><span class="pl-c1">1.745505</span>
<span class="pl-c1">2</span> <span class="pl-k">-</span><span class="pl-c1">0.263952</span> <span class="pl-c1">0.991460</span> <span class="pl-k">-</span><span class="pl-c1">0.919069</span> <span class="pl-c1">0.266046</span>
<span class="pl-c1">3</span> <span class="pl-k">-</span><span class="pl-c1">0.709661</span> <span class="pl-c1">1.669052</span> <span class="pl-c1">1.037882</span> <span class="pl-k">-</span><span class="pl-c1">1.705775</span>
<span class="pl-c1">4</span> <span class="pl-k">-</span><span class="pl-c1">0.919854</span> <span class="pl-k">-</span><span class="pl-c1">0.042379</span> <span class="pl-c1">1.247642</span> <span class="pl-k">-</span><span class="pl-c1">0.009920</span>
<span class="pl-c1">5</span> <span class="pl-c1">0.290213</span> <span class="pl-c1">0.495767</span> <span class="pl-c1">0.362949</span> <span class="pl-c1">1.548106</span>
<span class="pl-c1">6</span> <span class="pl-k">-</span><span class="pl-c1">1.131345</span> <span class="pl-k">-</span><span class="pl-c1">0.089329</span> <span class="pl-c1">0.337863</span> <span class="pl-k">-</span><span class="pl-c1">0.945867</span>
<span class="pl-c1">7</span> <span class="pl-k">-</span><span class="pl-c1">0.932132</span> <span class="pl-c1">1.956030</span> <span class="pl-c1">0.017587</span> <span class="pl-k">-</span><span class="pl-c1">0.016692</span>
<span class="pl-c1">8</span> <span class="pl-k">-</span><span class="pl-c1">0.575247</span> <span class="pl-c1">0.254161</span> <span class="pl-k">-</span><span class="pl-c1">1.143704</span> <span class="pl-c1">0.215897</span>
<span class="pl-c1">9</span> <span class="pl-c1">1.193555</span> <span class="pl-k">-</span><span class="pl-c1">0.077118</span> <span class="pl-k">-</span><span class="pl-c1">0.408530</span> <span class="pl-k">-</span><span class="pl-c1">0.862495</span>
<span class="pl-c"><span class="pl-c">#</span> break it into pieces</span>
In [<span class="pl-c1">75</span>]: pieces <span class="pl-k">=</span> [df[:<span class="pl-c1">3</span>], df[<span class="pl-c1">3</span>:<span class="pl-c1">7</span>], df[<span class="pl-c1">7</span>:]]
In [<span class="pl-c1">76</span>]: pd.concat(pieces)
Out[<span class="pl-c1">76</span>]:
<span class="pl-c1">0</span> <span class="pl-c1">1</span> <span class="pl-c1">2</span> <span class="pl-c1">3</span>
<span class="pl-c1">0</span> <span class="pl-k">-</span><span class="pl-c1">0.548702</span> <span class="pl-c1">1.467327</span> <span class="pl-k">-</span><span class="pl-c1">1.015962</span> <span class="pl-k">-</span><span class="pl-c1">0.483075</span>
<span class="pl-c1">1</span> <span class="pl-c1">1.637550</span> <span class="pl-k">-</span><span class="pl-c1">1.217659</span> <span class="pl-k">-</span><span class="pl-c1">0.291519</span> <span class="pl-k">-</span><span class="pl-c1">1.745505</span>
<span class="pl-c1">2</span> <span class="pl-k">-</span><span class="pl-c1">0.263952</span> <span class="pl-c1">0.991460</span> <span class="pl-k">-</span><span class="pl-c1">0.919069</span> <span class="pl-c1">0.266046</span>
<span class="pl-c1">3</span> <span class="pl-k">-</span><span class="pl-c1">0.709661</span> <span class="pl-c1">1.669052</span> <span class="pl-c1">1.037882</span> <span class="pl-k">-</span><span class="pl-c1">1.705775</span>
<span class="pl-c1">4</span> <span class="pl-k">-</span><span class="pl-c1">0.919854</span> <span class="pl-k">-</span><span class="pl-c1">0.042379</span> <span class="pl-c1">1.247642</span> <span class="pl-k">-</span><span class="pl-c1">0.009920</span>
<span class="pl-c1">5</span> <span class="pl-c1">0.290213</span> <span class="pl-c1">0.495767</span> <span class="pl-c1">0.362949</span> <span class="pl-c1">1.548106</span>
<span class="pl-c1">6</span> <span class="pl-k">-</span><span class="pl-c1">1.131345</span> <span class="pl-k">-</span><span class="pl-c1">0.089329</span> <span class="pl-c1">0.337863</span> <span class="pl-k">-</span><span class="pl-c1">0.945867</span>
<span class="pl-c1">7</span> <span class="pl-k">-</span><span class="pl-c1">0.932132</span> <span class="pl-c1">1.956030</span> <span class="pl-c1">0.017587</span> <span class="pl-k">-</span><span class="pl-c1">0.016692</span>
<span class="pl-c1">8</span> <span class="pl-k">-</span><span class="pl-c1">0.575247</span> <span class="pl-c1">0.254161</span> <span class="pl-k">-</span><span class="pl-c1">1.143704</span> <span class="pl-c1">0.215897</span>
<span class="pl-c1">9</span> <span class="pl-c1">1.193555</span> <span class="pl-k">-</span><span class="pl-c1">0.077118</span> <span class="pl-k">-</span><span class="pl-c1">0.408530</span> <span class="pl-k">-</span><span class="pl-c1">0.862495</span></pre></div>
<h2><code>Join</code></h2>
<p>类似于 SQL 类型的合并，具体请参阅：<a href="http://pandas.pydata.org/pandas-docs/stable/merging.html#merging-join" rel="nofollow">数据库风格的连接</a></p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">77</span>]: left <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">'</span>key<span class="pl-pds">'</span></span>: [<span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>], <span class="pl-s"><span class="pl-pds">'</span>lval<span class="pl-pds">'</span></span>: [<span class="pl-c1">1</span>, <span class="pl-c1">2</span>]})
In [<span class="pl-c1">78</span>]: right <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">'</span>key<span class="pl-pds">'</span></span>: [<span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>], <span class="pl-s"><span class="pl-pds">'</span>rval<span class="pl-pds">'</span></span>: [<span class="pl-c1">4</span>, <span class="pl-c1">5</span>]})
In [<span class="pl-c1">79</span>]: left
Out[<span class="pl-c1">79</span>]:
key lval
<span class="pl-c1">0</span> foo <span class="pl-c1">1</span>
<span class="pl-c1">1</span> foo <span class="pl-c1">2</span>
In [<span class="pl-c1">80</span>]: right
Out[<span class="pl-c1">80</span>]:
key rval
<span class="pl-c1">0</span> foo <span class="pl-c1">4</span>
<span class="pl-c1">1</span> foo <span class="pl-c1">5</span>
In [<span class="pl-c1">81</span>]: pd.merge(left, right, <span class="pl-v">on</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>key<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">81</span>]:
key lval rval
<span class="pl-c1">0</span> foo <span class="pl-c1">1</span> <span class="pl-c1">4</span>
<span class="pl-c1">1</span> foo <span class="pl-c1">1</span> <span class="pl-c1">5</span>
<span class="pl-c1">2</span> foo <span class="pl-c1">2</span> <span class="pl-c1">4</span>
<span class="pl-c1">3</span> foo <span class="pl-c1">2</span> <span class="pl-c1">5</span></pre></div>
<p>另一个例子：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">82</span>]: left <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">'</span>key<span class="pl-pds">'</span></span>: [<span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>], <span class="pl-s"><span class="pl-pds">'</span>lval<span class="pl-pds">'</span></span>: [<span class="pl-c1">1</span>, <span class="pl-c1">2</span>]})
In [<span class="pl-c1">83</span>]: right <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">'</span>key<span class="pl-pds">'</span></span>: [<span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>], <span class="pl-s"><span class="pl-pds">'</span>rval<span class="pl-pds">'</span></span>: [<span class="pl-c1">4</span>, <span class="pl-c1">5</span>]})
In [<span class="pl-c1">84</span>]: left
Out[<span class="pl-c1">84</span>]:
key lval
<span class="pl-c1">0</span> foo <span class="pl-c1">1</span>
<span class="pl-c1">1</span> bar <span class="pl-c1">2</span>
In [<span class="pl-c1">85</span>]: right
Out[<span class="pl-c1">85</span>]:
key rval
<span class="pl-c1">0</span> foo <span class="pl-c1">4</span>
<span class="pl-c1">1</span> bar <span class="pl-c1">5</span>
In [<span class="pl-c1">86</span>]: pd.merge(left, right, <span class="pl-v">on</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>key<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">86</span>]:
key lval rval
<span class="pl-c1">0</span> foo <span class="pl-c1">1</span> <span class="pl-c1">4</span>
<span class="pl-c1">1</span> bar <span class="pl-c1">2</span> <span class="pl-c1">5</span></pre></div>
<h2><code>Append</code></h2>
<p>将一行连接到一个<code>DataFrame</code>上，具体请参阅<a href="http://pandas.pydata.org/pandas-docs/stable/merging.html#merging-concatenation" rel="nofollow">附加</a>：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">87</span>]: df <span class="pl-k">=</span> pd.DataFrame(np.random.randn(<span class="pl-c1">8</span>, <span class="pl-c1">4</span>), <span class="pl-v">columns</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>])
In [<span class="pl-c1">88</span>]: df
Out[<span class="pl-c1">88</span>]:
A B C D
<span class="pl-c1">0</span> <span class="pl-c1">1.346061</span> <span class="pl-c1">1.511763</span> <span class="pl-c1">1.627081</span> <span class="pl-k">-</span><span class="pl-c1">0.990582</span>
<span class="pl-c1">1</span> <span class="pl-k">-</span><span class="pl-c1">0.441652</span> <span class="pl-c1">1.211526</span> <span class="pl-c1">0.268520</span> <span class="pl-c1">0.024580</span>
<span class="pl-c1">2</span> <span class="pl-k">-</span><span class="pl-c1">1.577585</span> <span class="pl-c1">0.396823</span> <span class="pl-k">-</span><span class="pl-c1">0.105381</span> <span class="pl-k">-</span><span class="pl-c1">0.532532</span>
<span class="pl-c1">3</span> <span class="pl-c1">1.453749</span> <span class="pl-c1">1.208843</span> <span class="pl-k">-</span><span class="pl-c1">0.080952</span> <span class="pl-k">-</span><span class="pl-c1">0.264610</span>
<span class="pl-c1">4</span> <span class="pl-k">-</span><span class="pl-c1">0.727965</span> <span class="pl-k">-</span><span class="pl-c1">0.589346</span> <span class="pl-c1">0.339969</span> <span class="pl-k">-</span><span class="pl-c1">0.693205</span>
<span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">0.339355</span> <span class="pl-c1">0.593616</span> <span class="pl-c1">0.884345</span> <span class="pl-c1">1.591431</span>
<span class="pl-c1">6</span> <span class="pl-c1">0.141809</span> <span class="pl-c1">0.220390</span> <span class="pl-c1">0.435589</span> <span class="pl-c1">0.192451</span>
<span class="pl-c1">7</span> <span class="pl-k">-</span><span class="pl-c1">0.096701</span> <span class="pl-c1">0.803351</span> <span class="pl-c1">1.715071</span> <span class="pl-k">-</span><span class="pl-c1">0.708758</span>
In [<span class="pl-c1">89</span>]: s <span class="pl-k">=</span> df.iloc[<span class="pl-c1">3</span>]
In [<span class="pl-c1">90</span>]: df.append(s, <span class="pl-v">ignore_index</span><span class="pl-k">=</span><span class="pl-c1">True</span>)
Out[<span class="pl-c1">90</span>]:
A B C D
<span class="pl-c1">0</span> <span class="pl-c1">1.346061</span> <span class="pl-c1">1.511763</span> <span class="pl-c1">1.627081</span> <span class="pl-k">-</span><span class="pl-c1">0.990582</span>
<span class="pl-c1">1</span> <span class="pl-k">-</span><span class="pl-c1">0.441652</span> <span class="pl-c1">1.211526</span> <span class="pl-c1">0.268520</span> <span class="pl-c1">0.024580</span>
<span class="pl-c1">2</span> <span class="pl-k">-</span><span class="pl-c1">1.577585</span> <span class="pl-c1">0.396823</span> <span class="pl-k">-</span><span class="pl-c1">0.105381</span> <span class="pl-k">-</span><span class="pl-c1">0.532532</span>
<span class="pl-c1">3</span> <span class="pl-c1">1.453749</span> <span class="pl-c1">1.208843</span> <span class="pl-k">-</span><span class="pl-c1">0.080952</span> <span class="pl-k">-</span><span class="pl-c1">0.264610</span>
<span class="pl-c1">4</span> <span class="pl-k">-</span><span class="pl-c1">0.727965</span> <span class="pl-k">-</span><span class="pl-c1">0.589346</span> <span class="pl-c1">0.339969</span> <span class="pl-k">-</span><span class="pl-c1">0.693205</span>
<span class="pl-c1">5</span> <span class="pl-k">-</span><span class="pl-c1">0.339355</span> <span class="pl-c1">0.593616</span> <span class="pl-c1">0.884345</span> <span class="pl-c1">1.591431</span>
<span class="pl-c1">6</span> <span class="pl-c1">0.141809</span> <span class="pl-c1">0.220390</span> <span class="pl-c1">0.435589</span> <span class="pl-c1">0.192451</span>
<span class="pl-c1">7</span> <span class="pl-k">-</span><span class="pl-c1">0.096701</span> <span class="pl-c1">0.803351</span> <span class="pl-c1">1.715071</span> <span class="pl-k">-</span><span class="pl-c1">0.708758</span>
<span class="pl-c1">8</span> <span class="pl-c1">1.453749</span> <span class="pl-c1">1.208843</span> <span class="pl-k">-</span><span class="pl-c1">0.080952</span> <span class="pl-k">-</span><span class="pl-c1">0.264610</span></pre></div>
<h1>七、 分组</h1>
<p>对于”group by”操作，我们通常是指以下一个或多个操作步骤：</p>
<ul>
<li>
<p>（Splitting）按照一些规则将数据分为不同的组；</p>
</li>
<li>
<p>（Applying）对于每组数据分别执行一个函数；</p>
</li>
<li>
<p>（Combining）将结果组合到一个数据结构中；</p>
</li>
</ul>
<p>详情请参阅：<a href="http://pandas.pydata.org/pandas-docs/stable/groupby.html#groupby" rel="nofollow"><em>Grouping section</em></a></p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">91</span>]: df <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span> : [<span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>,
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>],
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span> : [<span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>three<span class="pl-pds">'</span></span>,
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>three<span class="pl-pds">'</span></span>],
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span> : np.random.randn(<span class="pl-c1">8</span>),
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span> : np.random.randn(<span class="pl-c1">8</span>)})
<span class="pl-c1">...</span>.:
In [<span class="pl-c1">92</span>]: df
Out[<span class="pl-c1">92</span>]:
A B C D
<span class="pl-c1">0</span> foo one <span class="pl-k">-</span><span class="pl-c1">1.202872</span> <span class="pl-k">-</span><span class="pl-c1">0.055224</span>
<span class="pl-c1">1</span> bar one <span class="pl-k">-</span><span class="pl-c1">1.814470</span> <span class="pl-c1">2.395985</span>
<span class="pl-c1">2</span> foo two <span class="pl-c1">1.018601</span> <span class="pl-c1">1.552825</span>
<span class="pl-c1">3</span> bar three <span class="pl-k">-</span><span class="pl-c1">0.595447</span> <span class="pl-c1">0.166599</span>
<span class="pl-c1">4</span> foo two <span class="pl-c1">1.395433</span> <span class="pl-c1">0.047609</span>
<span class="pl-c1">5</span> bar two <span class="pl-k">-</span><span class="pl-c1">0.392670</span> <span class="pl-k">-</span><span class="pl-c1">0.136473</span>
<span class="pl-c1">6</span> foo one <span class="pl-c1">0.007207</span> <span class="pl-k">-</span><span class="pl-c1">0.561757</span>
<span class="pl-c1">7</span> foo three <span class="pl-c1">1.928123</span> <span class="pl-k">-</span><span class="pl-c1">1.623033</span></pre></div>
<p>1、 分组并对每个分组执行<code>sum</code>函数：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">93</span>]: df.groupby(<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>).sum()
Out[<span class="pl-c1">93</span>]:
C D
A
bar <span class="pl-k">-</span><span class="pl-c1">2.802588</span> <span class="pl-c1">2.42611</span>
foo <span class="pl-c1">3.146492</span> <span class="pl-k">-</span><span class="pl-c1">0.63958</span></pre></div>
<p>2、 通过多个列进行分组形成一个层次索引，然后执行函数：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">94</span>]: df.groupby([<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>]).sum()
Out[<span class="pl-c1">94</span>]:
C D
A B
bar one <span class="pl-k">-</span><span class="pl-c1">1.814470</span> <span class="pl-c1">2.395985</span>
three <span class="pl-k">-</span><span class="pl-c1">0.595447</span> <span class="pl-c1">0.166599</span>
two <span class="pl-k">-</span><span class="pl-c1">0.392670</span> <span class="pl-k">-</span><span class="pl-c1">0.136473</span>
foo one <span class="pl-k">-</span><span class="pl-c1">1.195665</span> <span class="pl-k">-</span><span class="pl-c1">0.616981</span>
three <span class="pl-c1">1.928123</span> <span class="pl-k">-</span><span class="pl-c1">1.623033</span>
two <span class="pl-c1">2.414034</span> <span class="pl-c1">1.600434</span></pre></div>
<h1>八、 改变形状</h1>
<p>详情请参阅 <a href="http://pandas.pydata.org/pandas-docs/stable/advanced.html#advanced-hierarchical" rel="nofollow">层次索引</a> 和 <a href="http://pandas.pydata.org/pandas-docs/stable/reshaping.html#reshaping-stacking" rel="nofollow">改变形状</a>。</p>
<h2><code>Stack</code></h2>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">95</span>]: tuples <span class="pl-k">=</span> <span class="pl-c1">list</span>(<span class="pl-c1">zip</span>(<span class="pl-k">*</span>[[<span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>baz<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>baz<span class="pl-pds">'</span></span>,
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>qux<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>qux<span class="pl-pds">'</span></span>],
<span class="pl-c1">...</span>.: [<span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>,
<span class="pl-c1">...</span>.: <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>]]))
<span class="pl-c1">...</span>.:
In [<span class="pl-c1">96</span>]: index <span class="pl-k">=</span> pd.MultiIndex.from_tuples(tuples, <span class="pl-v">names</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>first<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>second<span class="pl-pds">'</span></span>])
In [<span class="pl-c1">97</span>]: df <span class="pl-k">=</span> pd.DataFrame(np.random.randn(<span class="pl-c1">8</span>, <span class="pl-c1">2</span>), <span class="pl-v">index</span><span class="pl-k">=</span>index, <span class="pl-v">columns</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>])
In [<span class="pl-c1">98</span>]: df2 <span class="pl-k">=</span> df[:<span class="pl-c1">4</span>]
In [<span class="pl-c1">99</span>]: df2
Out[<span class="pl-c1">99</span>]:
A B
first second
bar one <span class="pl-c1">0.029399</span> <span class="pl-k">-</span><span class="pl-c1">0.542108</span>
two <span class="pl-c1">0.282696</span> <span class="pl-k">-</span><span class="pl-c1">0.087302</span>
baz one <span class="pl-k">-</span><span class="pl-c1">1.575170</span> <span class="pl-c1">1.771208</span>
two <span class="pl-c1">0.816482</span> <span class="pl-c1">1.100230</span></pre></div>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">100</span>]: stacked <span class="pl-k">=</span> df2.stack()
In [<span class="pl-c1">101</span>]: stacked
Out[<span class="pl-c1">101</span>]:
first second
bar one A <span class="pl-c1">0.029399</span>
B <span class="pl-k">-</span><span class="pl-c1">0.542108</span>
two A <span class="pl-c1">0.282696</span>
B <span class="pl-k">-</span><span class="pl-c1">0.087302</span>
baz one A <span class="pl-k">-</span><span class="pl-c1">1.575170</span>
B <span class="pl-c1">1.771208</span>
two A <span class="pl-c1">0.816482</span>
B <span class="pl-c1">1.100230</span>
dtype: float64</pre></div>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">102</span>]: stacked.unstack()
Out[<span class="pl-c1">102</span>]:
A B
first second
bar one <span class="pl-c1">0.029399</span> <span class="pl-k">-</span><span class="pl-c1">0.542108</span>
two <span class="pl-c1">0.282696</span> <span class="pl-k">-</span><span class="pl-c1">0.087302</span>
baz one <span class="pl-k">-</span><span class="pl-c1">1.575170</span> <span class="pl-c1">1.771208</span>
two <span class="pl-c1">0.816482</span> <span class="pl-c1">1.100230</span>
In [<span class="pl-c1">103</span>]: stacked.unstack(<span class="pl-c1">1</span>)
Out[<span class="pl-c1">103</span>]:
second one two
first
bar A <span class="pl-c1">0.029399</span> <span class="pl-c1">0.282696</span>
B <span class="pl-k">-</span><span class="pl-c1">0.542108</span> <span class="pl-k">-</span><span class="pl-c1">0.087302</span>
baz A <span class="pl-k">-</span><span class="pl-c1">1.575170</span> <span class="pl-c1">0.816482</span>
B <span class="pl-c1">1.771208</span> <span class="pl-c1">1.100230</span>
In [<span class="pl-c1">104</span>]: stacked.unstack(<span class="pl-c1">0</span>)
Out[<span class="pl-c1">104</span>]:
first bar baz
second
one A <span class="pl-c1">0.029399</span> <span class="pl-k">-</span><span class="pl-c1">1.575170</span>
B <span class="pl-k">-</span><span class="pl-c1">0.542108</span> <span class="pl-c1">1.771208</span>
two A <span class="pl-c1">0.282696</span> <span class="pl-c1">0.816482</span>
B <span class="pl-k">-</span><span class="pl-c1">0.087302</span> <span class="pl-c1">1.100230</span></pre></div>
<h2>数据透视表</h2>
<p>详情请参阅：<a href="http://pandas.pydata.org/pandas-docs/stable/reshaping.html#reshaping-pivot" rel="nofollow">数据透视表</a>.</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">105</span>]: df <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span> : [<span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>one<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>two<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>three<span class="pl-pds">'</span></span>] <span class="pl-k">*</span> <span class="pl-c1">3</span>,
<span class="pl-c1">...</span>..: <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span> : [<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span>] <span class="pl-k">*</span> <span class="pl-c1">4</span>,
<span class="pl-c1">...</span>..: <span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span> : [<span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>foo<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>bar<span class="pl-pds">'</span></span>] <span class="pl-k">*</span> <span class="pl-c1">2</span>,
<span class="pl-c1">...</span>..: <span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span> : np.random.randn(<span class="pl-c1">12</span>),
<span class="pl-c1">...</span>..: <span class="pl-s"><span class="pl-pds">'</span>E<span class="pl-pds">'</span></span> : np.random.randn(<span class="pl-c1">12</span>)})
<span class="pl-c1">...</span>..:
In [<span class="pl-c1">106</span>]: df
Out[<span class="pl-c1">106</span>]:
A B C D E
<span class="pl-c1">0</span> one A foo <span class="pl-c1">1.418757</span> <span class="pl-k">-</span><span class="pl-c1">0.179666</span>
<span class="pl-c1">1</span> one B foo <span class="pl-k">-</span><span class="pl-c1">1.879024</span> <span class="pl-c1">1.291836</span>
<span class="pl-c1">2</span> two C foo <span class="pl-c1">0.536826</span> <span class="pl-k">-</span><span class="pl-c1">0.009614</span>
<span class="pl-c1">3</span> three A bar <span class="pl-c1">1.006160</span> <span class="pl-c1">0.392149</span>
<span class="pl-c1">4</span> one B bar <span class="pl-k">-</span><span class="pl-c1">0.029716</span> <span class="pl-c1">0.264599</span>
<span class="pl-c1">5</span> one C bar <span class="pl-k">-</span><span class="pl-c1">1.146178</span> <span class="pl-k">-</span><span class="pl-c1">0.057409</span>
<span class="pl-c1">6</span> two A foo <span class="pl-c1">0.100900</span> <span class="pl-k">-</span><span class="pl-c1">1.425638</span>
<span class="pl-c1">7</span> three B foo <span class="pl-k">-</span><span class="pl-c1">1.035018</span> <span class="pl-c1">1.024098</span>
<span class="pl-c1">8</span> one C foo <span class="pl-c1">0.314665</span> <span class="pl-k">-</span><span class="pl-c1">0.106062</span>
<span class="pl-c1">9</span> one A bar <span class="pl-k">-</span><span class="pl-c1">0.773723</span> <span class="pl-c1">1.824375</span>
<span class="pl-c1">10</span> two B bar <span class="pl-k">-</span><span class="pl-c1">1.170653</span> <span class="pl-c1">0.595974</span>
<span class="pl-c1">11</span> three C bar <span class="pl-c1">0.648740</span> <span class="pl-c1">1.167115</span></pre></div>
<p>可以从这个数据中轻松的生成数据透视表：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">107</span>]: pd.pivot_table(df, <span class="pl-v">values</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>, <span class="pl-v">index</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>], <span class="pl-v">columns</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span>])
Out[<span class="pl-c1">107</span>]:
C bar foo
A B
one A <span class="pl-k">-</span><span class="pl-c1">0.773723</span> <span class="pl-c1">1.418757</span>
B <span class="pl-k">-</span><span class="pl-c1">0.029716</span> <span class="pl-k">-</span><span class="pl-c1">1.879024</span>
C <span class="pl-k">-</span><span class="pl-c1">1.146178</span> <span class="pl-c1">0.314665</span>
three A <span class="pl-c1">1.006160</span> NaN
B NaN <span class="pl-k">-</span><span class="pl-c1">1.035018</span>
C <span class="pl-c1">0.648740</span> NaN
two A NaN <span class="pl-c1">0.100900</span>
B <span class="pl-k">-</span><span class="pl-c1">1.170653</span> NaN
C NaN <span class="pl-c1">0.536826</span></pre></div>
<h1>九、 时间序列</h1>
<p>Pandas 在对频率转换进行重新采样时拥有简单、强大且高效的功能（如将按秒采样的数据转换为按5分钟为单位进行采样的数据）。这种操作在金融领域非常常见。具体参考：<a href="http://pandas.pydata.org/pandas-docs/stable/timeseries.html#timeseries" rel="nofollow">时间序列</a>。</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">108</span>]: rng <span class="pl-k">=</span> pd.date_range(<span class="pl-s"><span class="pl-pds">'</span>1/1/2012<span class="pl-pds">'</span></span>, <span class="pl-v">periods</span><span class="pl-k">=</span><span class="pl-c1">100</span>, <span class="pl-v">freq</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>S<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">109</span>]: ts <span class="pl-k">=</span> pd.Series(np.random.randint(<span class="pl-c1">0</span>, <span class="pl-c1">500</span>, <span class="pl-c1">len</span>(rng)), <span class="pl-v">index</span><span class="pl-k">=</span>rng)
In [<span class="pl-c1">110</span>]: ts.resample(<span class="pl-s"><span class="pl-pds">'</span>5Min<span class="pl-pds">'</span></span>).sum()
Out[<span class="pl-c1">110</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">25083</span>
Freq: <span class="pl-ii">5T</span>, dtype: int64</pre></div>
<p>1、 时区表示：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">111</span>]: rng <span class="pl-k">=</span> pd.date_range(<span class="pl-s"><span class="pl-pds">'</span>3/6/2012 00:00<span class="pl-pds">'</span></span>, <span class="pl-v">periods</span><span class="pl-k">=</span><span class="pl-c1">5</span>, <span class="pl-v">freq</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">112</span>]: ts <span class="pl-k">=</span> pd.Series(np.random.randn(<span class="pl-c1">len</span>(rng)), rng)
In [<span class="pl-c1">113</span>]: ts
Out[<span class="pl-c1">113</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">0.464000</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">0.227371</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">8</span></span> <span class="pl-k">-</span><span class="pl-c1">0.496922</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span> <span class="pl-c1">0.306389</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">10</span> <span class="pl-k">-</span><span class="pl-c1">2.290613</span>
Freq: D, dtype: float64
In [<span class="pl-c1">114</span>]: ts_utc <span class="pl-k">=</span> ts.tz_localize(<span class="pl-s"><span class="pl-pds">'</span>UTC<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">115</span>]: ts_utc
Out[<span class="pl-c1">115</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">+</span><span class="pl-c1">00</span>:<span class="pl-c1">00</span> <span class="pl-c1">0.464000</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">+</span><span class="pl-c1">00</span>:<span class="pl-c1">00</span> <span class="pl-c1">0.227371</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">8</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">+</span><span class="pl-c1">00</span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">0.496922</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">+</span><span class="pl-c1">00</span>:<span class="pl-c1">00</span> <span class="pl-c1">0.306389</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">10</span> <span class="pl-c1">00</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">+</span><span class="pl-c1">00</span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">2.290613</span>
Freq: D, dtype: float64</pre></div>
<p>2、 时区转换：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">116</span>]: ts_utc.tz_convert(<span class="pl-s"><span class="pl-pds">'</span>US/Eastern<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">116</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">19</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span>:<span class="pl-c1">00</span> <span class="pl-c1">0.464000</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">19</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span>:<span class="pl-c1">00</span> <span class="pl-c1">0.227371</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">19</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">0.496922</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">8</span></span> <span class="pl-c1">19</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span>:<span class="pl-c1">00</span> <span class="pl-c1">0.306389</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span> <span class="pl-c1">19</span>:<span class="pl-c1">00</span>:<span class="pl-c1">00</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">2.290613</span>
Freq: D, dtype: float64</pre></div>
<p>3、 时间跨度转换：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">117</span>]: rng <span class="pl-k">=</span> pd.date_range(<span class="pl-s"><span class="pl-pds">'</span>1/1/2012<span class="pl-pds">'</span></span>, <span class="pl-v">periods</span><span class="pl-k">=</span><span class="pl-c1">5</span>, <span class="pl-v">freq</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>M<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">118</span>]: ts <span class="pl-k">=</span> pd.Series(np.random.randn(<span class="pl-c1">len</span>(rng)), <span class="pl-v">index</span><span class="pl-k">=</span>rng)
In [<span class="pl-c1">119</span>]: ts
Out[<span class="pl-c1">119</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">31</span> <span class="pl-k">-</span><span class="pl-c1">1.134623</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span><span class="pl-k">-</span><span class="pl-c1">29</span> <span class="pl-k">-</span><span class="pl-c1">1.561819</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">31</span> <span class="pl-k">-</span><span class="pl-c1">0.260838</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span><span class="pl-k">-</span><span class="pl-c1">30</span> <span class="pl-c1">0.281957</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span><span class="pl-k">-</span><span class="pl-c1">31</span> <span class="pl-c1">1.523962</span>
Freq: M, dtype: float64
In [<span class="pl-c1">120</span>]: ps <span class="pl-k">=</span> ts.to_period()
In [<span class="pl-c1">121</span>]: ps
Out[<span class="pl-c1">121</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-k">-</span><span class="pl-c1">1.134623</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">1.561819</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">0.260838</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-c1">0.281957</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">1.523962</span>
Freq: M, dtype: float64
In [<span class="pl-c1">122</span>]: ps.to_timestamp()
Out[<span class="pl-c1">122</span>]:
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-k">-</span><span class="pl-c1">1.134623</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-k">-</span><span class="pl-c1">1.561819</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-k">-</span><span class="pl-c1">0.260838</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.281957</span>
<span class="pl-c1">2012</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">1.523962</span>
Freq: <span class="pl-c1">MS</span>, dtype: float64</pre></div>
<p>4、 时期和时间戳之间的转换使得可以使用一些方便的算术函数。</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">123</span>]: prng <span class="pl-k">=</span> pd.period_range(<span class="pl-s"><span class="pl-pds">'</span>1990Q1<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>2000Q4<span class="pl-pds">'</span></span>, <span class="pl-v">freq</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>Q-NOV<span class="pl-pds">'</span></span>)
In [<span class="pl-c1">124</span>]: ts <span class="pl-k">=</span> pd.Series(np.random.randn(<span class="pl-c1">len</span>(prng)), prng)
In [<span class="pl-c1">125</span>]: ts.index <span class="pl-k">=</span> (prng.asfreq(<span class="pl-s"><span class="pl-pds">'</span>M<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>e<span class="pl-pds">'</span></span>) <span class="pl-k">+</span> <span class="pl-c1">1</span>).asfreq(<span class="pl-s"><span class="pl-pds">'</span>H<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>s<span class="pl-pds">'</span></span>) <span class="pl-k">+</span> <span class="pl-c1">9</span>
In [<span class="pl-c1">126</span>]: ts.head()
Out[<span class="pl-c1">126</span>]:
<span class="pl-c1">1990</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0<span class="pl-ii">9</span></span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">0.902937</span>
<span class="pl-c1">1990</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0<span class="pl-ii">9</span></span>:<span class="pl-c1">00</span> <span class="pl-c1">0.068159</span>
<span class="pl-c1">1990</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0<span class="pl-ii">9</span></span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">0.057873</span>
<span class="pl-c1">1990</span><span class="pl-k">-</span><span class="pl-c1">12</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0<span class="pl-ii">9</span></span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">0.368204</span>
<span class="pl-c1">1991</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0<span class="pl-ii">9</span></span>:<span class="pl-c1">00</span> <span class="pl-k">-</span><span class="pl-c1">1.144073</span>
Freq: H, dtype: float64</pre></div>
<h1>十、 Categorical</h1>
<p>从 0.15 版本开始，pandas 可以在<code>DataFrame</code>中支持 Categorical 类型的数据，详细 介绍参看：<a href="http://pandas.pydata.org/pandas-docs/stable/categorical.html#categorical" rel="nofollow">Categorical 简介</a>和<a href="http://pandas.pydata.org/pandas-docs/stable/api.html#api-categorical" rel="nofollow"><em>API documentation</em></a>。</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">127</span>]: df <span class="pl-k">=</span> pd.DataFrame({<span class="pl-s"><span class="pl-pds">"</span>id<span class="pl-pds">"</span></span>:[<span class="pl-c1">1</span>,<span class="pl-c1">2</span>,<span class="pl-c1">3</span>,<span class="pl-c1">4</span>,<span class="pl-c1">5</span>,<span class="pl-c1">6</span>], <span class="pl-s"><span class="pl-pds">"</span>raw_grade<span class="pl-pds">"</span></span>:[<span class="pl-s"><span class="pl-pds">'</span>a<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>b<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>b<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>a<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>a<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>e<span class="pl-pds">'</span></span>]})</pre></div>
<p>1、 将原始的<code>grade</code>转换为 Categorical 数据类型：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">128</span>]: df[<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>] <span class="pl-k">=</span> df[<span class="pl-s"><span class="pl-pds">"</span>raw_grade<span class="pl-pds">"</span></span>].astype(<span class="pl-s"><span class="pl-pds">"</span>category<span class="pl-pds">"</span></span>)
In [<span class="pl-c1">129</span>]: df[<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>]
Out[<span class="pl-c1">129</span>]:
<span class="pl-c1">0</span> a
<span class="pl-c1">1</span> b
<span class="pl-c1">2</span> b
<span class="pl-c1">3</span> a
<span class="pl-c1">4</span> a
<span class="pl-c1">5</span> e
Name: grade, dtype: category
Categories (<span class="pl-c1">3</span>, <span class="pl-c1">object</span>): [a, b, e]</pre></div>
<p>2、 将 Categorical 类型数据重命名为更有意义的名称：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">130</span>]: df[<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>].cat.categories <span class="pl-k">=</span> [<span class="pl-s"><span class="pl-pds">"</span>very good<span class="pl-pds">"</span></span>, <span class="pl-s"><span class="pl-pds">"</span>good<span class="pl-pds">"</span></span>, <span class="pl-s"><span class="pl-pds">"</span>very bad<span class="pl-pds">"</span></span>]</pre></div>
<p>3、 对类别进行重新排序，增加缺失的类别：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">131</span>]: df[<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>] <span class="pl-k">=</span> df[<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>].cat.set_categories([<span class="pl-s"><span class="pl-pds">"</span>very bad<span class="pl-pds">"</span></span>, <span class="pl-s"><span class="pl-pds">"</span>bad<span class="pl-pds">"</span></span>, <span class="pl-s"><span class="pl-pds">"</span>medium<span class="pl-pds">"</span></span>, <span class="pl-s"><span class="pl-pds">"</span>good<span class="pl-pds">"</span></span>, <span class="pl-s"><span class="pl-pds">"</span>very good<span class="pl-pds">"</span></span>])
In [<span class="pl-c1">132</span>]: df[<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>]
Out[<span class="pl-c1">132</span>]:
<span class="pl-c1">0</span> very good
<span class="pl-c1">1</span> good
<span class="pl-c1">2</span> good
<span class="pl-c1">3</span> very good
<span class="pl-c1">4</span> very good
<span class="pl-c1">5</span> very bad
Name: grade, dtype: category
Categories (<span class="pl-c1">5</span>, <span class="pl-c1">object</span>): [very bad, bad, medium, good, very good]</pre></div>
<p>4、 排序是按照 Categorical 的顺序进行的而不是按照字典顺序进行：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">133</span>]: df.sort_values(<span class="pl-v">by</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>)
Out[<span class="pl-c1">133</span>]:
<span class="pl-c1">id</span> raw_grade grade
<span class="pl-c1">5</span> <span class="pl-c1">6</span> e very bad
<span class="pl-c1">1</span> <span class="pl-c1">2</span> b good
<span class="pl-c1">2</span> <span class="pl-c1">3</span> b good
<span class="pl-c1">0</span> <span class="pl-c1">1</span> a very good
<span class="pl-c1">3</span> <span class="pl-c1">4</span> a very good
<span class="pl-c1">4</span> <span class="pl-c1">5</span> a very good</pre></div>
<p>5、 对 Categorical 列进行排序时存在空的类别：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">134</span>]: df.groupby(<span class="pl-s"><span class="pl-pds">"</span>grade<span class="pl-pds">"</span></span>).size()
Out[<span class="pl-c1">134</span>]:
grade
very bad <span class="pl-c1">1</span>
bad <span class="pl-c1">0</span>
medium <span class="pl-c1">0</span>
good <span class="pl-c1">2</span>
very good <span class="pl-c1">3</span>
dtype: int64</pre></div>
<h1>十一、 画图</h1>
<p>具体文档参看：<a href="http://pandas.pydata.org/pandas-docs/stable/visualization.html#visualization" rel="nofollow">绘图</a>文档。</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">135</span>]: ts <span class="pl-k">=</span> pd.Series(np.random.randn(<span class="pl-c1">1000</span>), <span class="pl-v">index</span><span class="pl-k">=</span>pd.date_range(<span class="pl-s"><span class="pl-pds">'</span>1/1/2000<span class="pl-pds">'</span></span>, <span class="pl-v">periods</span><span class="pl-k">=</span><span class="pl-c1">1000</span>))
In [<span class="pl-c1">136</span>]: ts <span class="pl-k">=</span> ts.cumsum()
In [<span class="pl-c1">137</span>]: ts.plot()
Out[<span class="pl-c1">137</span>]: <span class="pl-k"><</span>matplotlib.axes._subplots.AxesSubplot at <span class="pl-c1"><span class="pl-k">0x</span>7ff2ab2af550</span><span class="pl-k">></span></pre></div>
<p><a href="https://camo.githubusercontent.com/5cb54847c6c4bfe86192deed38660984d387500f/687474703a2f2f70616e6461732e7079646174612e6f72672f70616e6461732d646f63732f737461626c652f5f696d616765732f7365726965735f706c6f745f62617369632e706e67" target="_blank"><img src="https://camo.githubusercontent.com/5cb54847c6c4bfe86192deed38660984d387500f/687474703a2f2f70616e6461732e7079646174612e6f72672f70616e6461732d646f63732f737461626c652f5f696d616765732f7365726965735f706c6f745f62617369632e706e67" alt="" data-canonical-src="http://pandas.pydata.org/pandas-docs/stable/_images/series_plot_basic.png" style="max-width:100%;"></a></p>
<p>对于<code>DataFrame</code>来说，<code>plot</code>是一种将所有列及其标签进行绘制的简便方法：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">138</span>]: df <span class="pl-k">=</span> pd.DataFrame(np.random.randn(<span class="pl-c1">1000</span>, <span class="pl-c1">4</span>), <span class="pl-v">index</span><span class="pl-k">=</span>ts.index,
<span class="pl-c1">...</span>..: <span class="pl-v">columns</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>A<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>B<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>C<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>D<span class="pl-pds">'</span></span>])
<span class="pl-c1">...</span>..:
In [<span class="pl-c1">139</span>]: df <span class="pl-k">=</span> df.cumsum()
In [<span class="pl-c1">140</span>]: plt.figure(); df.plot(); plt.legend(<span class="pl-v">loc</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>best<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">140</span>]: <span class="pl-k"><</span>matplotlib.legend.Legend at <span class="pl-c1"><span class="pl-k">0x</span>7ff29c8163d0</span><span class="pl-k">></span></pre></div>
<p><a href="https://camo.githubusercontent.com/5185a50572882b99063347837e4f962286ef4d02/687474703a2f2f70616e6461732e7079646174612e6f72672f70616e6461732d646f63732f737461626c652f5f696d616765732f6672616d655f706c6f745f62617369632e706e67" target="_blank"><img src="https://camo.githubusercontent.com/5185a50572882b99063347837e4f962286ef4d02/687474703a2f2f70616e6461732e7079646174612e6f72672f70616e6461732d646f63732f737461626c652f5f696d616765732f6672616d655f706c6f745f62617369632e706e67" alt="" data-canonical-src="http://pandas.pydata.org/pandas-docs/stable/_images/frame_plot_basic.png" style="max-width:100%;"></a></p>
<h1>十二、 导入和保存数据</h1>
<h2>CSV</h2>
<p>参考：<a href="http://pandas.pydata.org/pandas-docs/stable/io.html#io-store-in-csv" rel="nofollow">写入 CSV 文件</a>。</p>
<p>1、 写入 csv 文件：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">141</span>]: df.to_csv(<span class="pl-s"><span class="pl-pds">'</span>foo.csv<span class="pl-pds">'</span></span>)</pre></div>
<p>2、 从 csv 文件中读取：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">142</span>]: pd.read_csv(<span class="pl-s"><span class="pl-pds">'</span>foo.csv<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">142</span>]:
Unnamed: <span class="pl-c1">0</span> A B C D
<span class="pl-c1">0</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.266457</span> <span class="pl-k">-</span><span class="pl-c1">0.399641</span> <span class="pl-k">-</span><span class="pl-c1">0.219582</span> <span class="pl-c1">1.186860</span>
<span class="pl-c1">1</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">1.170732</span> <span class="pl-k">-</span><span class="pl-c1">0.345873</span> <span class="pl-c1">1.653061</span> <span class="pl-k">-</span><span class="pl-c1">0.282953</span>
<span class="pl-c1">2</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">1.734933</span> <span class="pl-c1">0.530468</span> <span class="pl-c1">2.060811</span> <span class="pl-k">-</span><span class="pl-c1">0.515536</span>
<span class="pl-c1">3</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-k">-</span><span class="pl-c1">1.555121</span> <span class="pl-c1">1.452620</span> <span class="pl-c1">0.239859</span> <span class="pl-k">-</span><span class="pl-c1">1.156896</span>
<span class="pl-c1">4</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">0.578117</span> <span class="pl-c1">0.511371</span> <span class="pl-c1">0.103552</span> <span class="pl-k">-</span><span class="pl-c1">2.428202</span>
<span class="pl-c1">5</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">0.478344</span> <span class="pl-c1">0.449933</span> <span class="pl-k">-</span><span class="pl-c1">0.741620</span> <span class="pl-k">-</span><span class="pl-c1">1.962409</span>
<span class="pl-c1">6</span> <span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">1.235339</span> <span class="pl-k">-</span><span class="pl-c1">0.091757</span> <span class="pl-k">-</span><span class="pl-c1">1.543861</span> <span class="pl-k">-</span><span class="pl-c1">1.084753</span>
.. <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span>
<span class="pl-c1">993</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">20</span> <span class="pl-k">-</span><span class="pl-c1">10.628548</span> <span class="pl-k">-</span><span class="pl-c1">9.153563</span> <span class="pl-k">-</span><span class="pl-c1">7.883146</span> <span class="pl-c1">28.313940</span>
<span class="pl-c1">994</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">21</span> <span class="pl-k">-</span><span class="pl-c1">10.390377</span> <span class="pl-k">-</span><span class="pl-c1">8.727491</span> <span class="pl-k">-</span><span class="pl-c1">6.399645</span> <span class="pl-c1">30.914107</span>
<span class="pl-c1">995</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">22</span> <span class="pl-k">-</span><span class="pl-c1">8.985362</span> <span class="pl-k">-</span><span class="pl-c1">8.485624</span> <span class="pl-k">-</span><span class="pl-c1">4.669462</span> <span class="pl-c1">31.367740</span>
<span class="pl-c1">996</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">23</span> <span class="pl-k">-</span><span class="pl-c1">9.558560</span> <span class="pl-k">-</span><span class="pl-c1">8.781216</span> <span class="pl-k">-</span><span class="pl-c1">4.499815</span> <span class="pl-c1">30.518439</span>
<span class="pl-c1">997</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">24</span> <span class="pl-k">-</span><span class="pl-c1">9.902058</span> <span class="pl-k">-</span><span class="pl-c1">9.340490</span> <span class="pl-k">-</span><span class="pl-c1">4.386639</span> <span class="pl-c1">30.105593</span>
<span class="pl-c1">998</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">25</span> <span class="pl-k">-</span><span class="pl-c1">10.216020</span> <span class="pl-k">-</span><span class="pl-c1">9.480682</span> <span class="pl-k">-</span><span class="pl-c1">3.933802</span> <span class="pl-c1">29.758560</span>
<span class="pl-c1">999</span> <span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">26</span> <span class="pl-k">-</span><span class="pl-c1">11.856774</span> <span class="pl-k">-</span><span class="pl-c1">10.671012</span> <span class="pl-k">-</span><span class="pl-c1">3.216025</span> <span class="pl-c1">29.369368</span>
[<span class="pl-c1">1000</span> rows x <span class="pl-c1">5</span> columns]</pre></div>
<h2>HDF5</h2>
<p>参考：<a href="http://pandas.pydata.org/pandas-docs/stable/io.html#io-hdf5" rel="nofollow">HDF5 存储</a></p>
<p>1、 写入 HDF5 存储：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">143</span>]: df.to_hdf(<span class="pl-s"><span class="pl-pds">'</span>foo.h5<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>df<span class="pl-pds">'</span></span>)</pre></div>
<p>2、 从 HDF5 存储中读取：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">144</span>]: pd.read_hdf(<span class="pl-s"><span class="pl-pds">'</span>foo.h5<span class="pl-pds">'</span></span>,<span class="pl-s"><span class="pl-pds">'</span>df<span class="pl-pds">'</span></span>)
Out[<span class="pl-c1">144</span>]:
A B C D
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.266457</span> <span class="pl-k">-</span><span class="pl-c1">0.399641</span> <span class="pl-k">-</span><span class="pl-c1">0.219582</span> <span class="pl-c1">1.186860</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">1.170732</span> <span class="pl-k">-</span><span class="pl-c1">0.345873</span> <span class="pl-c1">1.653061</span> <span class="pl-k">-</span><span class="pl-c1">0.282953</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">1.734933</span> <span class="pl-c1">0.530468</span> <span class="pl-c1">2.060811</span> <span class="pl-k">-</span><span class="pl-c1">0.515536</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-k">-</span><span class="pl-c1">1.555121</span> <span class="pl-c1">1.452620</span> <span class="pl-c1">0.239859</span> <span class="pl-k">-</span><span class="pl-c1">1.156896</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">0.578117</span> <span class="pl-c1">0.511371</span> <span class="pl-c1">0.103552</span> <span class="pl-k">-</span><span class="pl-c1">2.428202</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">0.478344</span> <span class="pl-c1">0.449933</span> <span class="pl-k">-</span><span class="pl-c1">0.741620</span> <span class="pl-k">-</span><span class="pl-c1">1.962409</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">1.235339</span> <span class="pl-k">-</span><span class="pl-c1">0.091757</span> <span class="pl-k">-</span><span class="pl-c1">1.543861</span> <span class="pl-k">-</span><span class="pl-c1">1.084753</span>
<span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">20</span> <span class="pl-k">-</span><span class="pl-c1">10.628548</span> <span class="pl-k">-</span><span class="pl-c1">9.153563</span> <span class="pl-k">-</span><span class="pl-c1">7.883146</span> <span class="pl-c1">28.313940</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">21</span> <span class="pl-k">-</span><span class="pl-c1">10.390377</span> <span class="pl-k">-</span><span class="pl-c1">8.727491</span> <span class="pl-k">-</span><span class="pl-c1">6.399645</span> <span class="pl-c1">30.914107</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">22</span> <span class="pl-k">-</span><span class="pl-c1">8.985362</span> <span class="pl-k">-</span><span class="pl-c1">8.485624</span> <span class="pl-k">-</span><span class="pl-c1">4.669462</span> <span class="pl-c1">31.367740</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">23</span> <span class="pl-k">-</span><span class="pl-c1">9.558560</span> <span class="pl-k">-</span><span class="pl-c1">8.781216</span> <span class="pl-k">-</span><span class="pl-c1">4.499815</span> <span class="pl-c1">30.518439</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">24</span> <span class="pl-k">-</span><span class="pl-c1">9.902058</span> <span class="pl-k">-</span><span class="pl-c1">9.340490</span> <span class="pl-k">-</span><span class="pl-c1">4.386639</span> <span class="pl-c1">30.105593</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">25</span> <span class="pl-k">-</span><span class="pl-c1">10.216020</span> <span class="pl-k">-</span><span class="pl-c1">9.480682</span> <span class="pl-k">-</span><span class="pl-c1">3.933802</span> <span class="pl-c1">29.758560</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">26</span> <span class="pl-k">-</span><span class="pl-c1">11.856774</span> <span class="pl-k">-</span><span class="pl-c1">10.671012</span> <span class="pl-k">-</span><span class="pl-c1">3.216025</span> <span class="pl-c1">29.369368</span>
[<span class="pl-c1">1000</span> rows x <span class="pl-c1">4</span> columns]</pre></div>
<h2>Excel</h2>
<p>参考：<a href="http://pandas.pydata.org/pandas-docs/stable/io.html#io-excel" rel="nofollow"><em>MS Excel</em></a></p>
<p>1、 写入excel文件：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">145</span>]: df.to_excel(<span class="pl-s"><span class="pl-pds">'</span>foo.xlsx<span class="pl-pds">'</span></span>, <span class="pl-v">sheet_name</span><span class="pl-k">=</span><span class="pl-s"><span class="pl-pds">'</span>Sheet1<span class="pl-pds">'</span></span>)</pre></div>
<p>2、 从excel文件中读取：</p>
<div class="highlight highlight-source-python"><pre>In [<span class="pl-c1">146</span>]: pd.read_excel(<span class="pl-s"><span class="pl-pds">'</span>foo.xlsx<span class="pl-pds">'</span></span>, <span class="pl-s"><span class="pl-pds">'</span>Sheet1<span class="pl-pds">'</span></span>, <span class="pl-v">index_col</span><span class="pl-k">=</span><span class="pl-c1">None</span>, <span class="pl-v">na_values</span><span class="pl-k">=</span>[<span class="pl-s"><span class="pl-pds">'</span>NA<span class="pl-pds">'</span></span>])
Out[<span class="pl-c1">146</span>]:
A B C D
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span> <span class="pl-c1">0.266457</span> <span class="pl-k">-</span><span class="pl-c1">0.399641</span> <span class="pl-k">-</span><span class="pl-c1">0.219582</span> <span class="pl-c1">1.186860</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">2</span></span> <span class="pl-k">-</span><span class="pl-c1">1.170732</span> <span class="pl-k">-</span><span class="pl-c1">0.345873</span> <span class="pl-c1">1.653061</span> <span class="pl-k">-</span><span class="pl-c1">0.282953</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">3</span></span> <span class="pl-k">-</span><span class="pl-c1">1.734933</span> <span class="pl-c1">0.530468</span> <span class="pl-c1">2.060811</span> <span class="pl-k">-</span><span class="pl-c1">0.515536</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">4</span></span> <span class="pl-k">-</span><span class="pl-c1">1.555121</span> <span class="pl-c1">1.452620</span> <span class="pl-c1">0.239859</span> <span class="pl-k">-</span><span class="pl-c1">1.156896</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">5</span></span> <span class="pl-c1">0.578117</span> <span class="pl-c1">0.511371</span> <span class="pl-c1">0.103552</span> <span class="pl-k">-</span><span class="pl-c1">2.428202</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">6</span></span> <span class="pl-c1">0.478344</span> <span class="pl-c1">0.449933</span> <span class="pl-k">-</span><span class="pl-c1">0.741620</span> <span class="pl-k">-</span><span class="pl-c1">1.962409</span>
<span class="pl-c1">2000</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">1</span></span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">7</span></span> <span class="pl-c1">1.235339</span> <span class="pl-k">-</span><span class="pl-c1">0.091757</span> <span class="pl-k">-</span><span class="pl-c1">1.543861</span> <span class="pl-k">-</span><span class="pl-c1">1.084753</span>
<span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span> <span class="pl-c1">...</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">20</span> <span class="pl-k">-</span><span class="pl-c1">10.628548</span> <span class="pl-k">-</span><span class="pl-c1">9.153563</span> <span class="pl-k">-</span><span class="pl-c1">7.883146</span> <span class="pl-c1">28.313940</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">21</span> <span class="pl-k">-</span><span class="pl-c1">10.390377</span> <span class="pl-k">-</span><span class="pl-c1">8.727491</span> <span class="pl-k">-</span><span class="pl-c1">6.399645</span> <span class="pl-c1">30.914107</span>
<span class="pl-c1">2002</span><span class="pl-k">-</span><span class="pl-c1">0<span class="pl-ii">9</span></span><span class="pl-k">-</span><span class="pl-c1">22</span> <span class="pl-k">-</span><span class="pl-c1">8.985362</span> <span class="pl-k">-</span><span class="pl-c1">8.485624</span> <span class="pl-k">-</span><span class="pl-c1">4.669462</span> <span class="pl-c1">31.367740</span>