python - 与重复索引合并 - 行数大于预期

Question

我有两个带有一些重复索引的数据框

df1 = pd.DataFrame(np.random.randn(5, 3), columns=['A', 'B', 'C'], index=['I1', 'I1' ,'I1', 'I2', 'I2'])
df2 = pd.DataFrame(np.random.randn(4, 3), columns=['D', 'E', 'F'], index=['I1', 'I1', 'I1', 'I2'])

pd.merge(df1, df2, how='left', left_index=True, right_index=True)

pandas 似乎无法识别两个索引具有相同的值。我预计数据框必须有 5 行，列为 A B C D E F。最后一行全部为 D E F

的 nans

类似这样的事情:

        A         B         C         D         E         F
I1  0.121993  0.208368 -0.056375  0.492218 -0.915034  1.667015
I1  0.121993  0.208368 -0.056375 -0.055575 -0.207215 -0.351027
I1  0.121993  0.208368 -0.056375  1.128143  1.371022  0.810542
I2 -0.817558  1.599293 -0.342841 -0.831796 -0.118316 -0.138027
I2 -0.817558  1.599293 -0.342841  NaN       NaN       NaN

相反，我得到了这个:

          A         B         C         D         E         F
I1  0.121993  0.208368 -0.056375  0.492218 -0.915034  1.667015
I1  0.121993  0.208368 -0.056375 -0.055575 -0.207215 -0.351027
I1  0.121993  0.208368 -0.056375  1.128143  1.371022  0.810542
I1  0.403085  0.532958  0.856544  0.492218 -0.915034  1.667015
I1  0.403085  0.532958  0.856544 -0.055575 -0.207215 -0.351027
I1  0.403085  0.532958  0.856544  1.128143  1.371022  0.810542
I1  0.094214 -0.527932 -1.368606  0.492218 -0.915034  1.667015
I1  0.094214 -0.527932 -1.368606 -0.055575 -0.207215 -0.351027
I1  0.094214 -0.527932 -1.368606  1.128143  1.371022  0.810542
I2  0.378565  0.331995  0.167682 -0.831796 -0.118316 -0.138027
I2  0.378565  0.331995  0.167682 -0.561473 -0.898151 -0.217683
I2 -0.817558  1.599293 -0.342841 -0.831796 -0.118316 -0.138027
I2 -0.817558  1.599293 -0.342841 -0.561473 -0.898151 -0.217683

编辑:我无法对结果进行重复数据删除，因为我不想丢失原始数据帧中的重复行

Answer 1

使用pd.concat()加入同一索引:

pd.concat([df2,df1],axis=1)

           A         B         C         D         E         F
I1  0.112906 -1.080809  0.857712 -0.849395  0.015475  0.619177
I1 -0.380070  1.389495  1.372172 -0.472603 -0.593138 -0.594146
I1 -0.258423  1.402873 -0.923191 -2.138440  0.099878  0.148920
I2 -1.618755 -0.459908 -0.803290 -0.267760  0.275084  0.810870
I2 -0.033210  0.523840 -1.028478 -1.300269 -1.516137  0.373555

编辑:

pd.concat([df1.reset_index(drop=True), df2.reset_index(drop=True)],axis=1).set_index(df1.index)

           A         B         C         D         E         F
I1  1.925637  0.082031  0.483414 -0.189940  0.763408 -0.346046
I1 -0.676511  0.482327  1.648381  2.635290 -0.080474  0.558633
I1  0.180004 -0.190909  0.821891 -1.010627  0.774914  0.988356
I2 -0.011089  0.364400 -0.207062 -1.335626  0.036884  1.628115
I2 -1.314910  0.294986  0.334418       NaN       NaN       NaN