pandas 内部合并/联接返回所有行 [英] Pandas inner merge/join returning all rows

问题描述

我试图基于两个字段中存在的列合并两个数据帧，仅保留两个集合的交集.

I'm trying to merge two data frames based on a column present in both, keeping only the intersection of the two sets.

所需的结果是:

 foo           bar            foobar
x y z        x j i           x y z j i
a 1 2        a 9 0           a 1 2 9 0
b 3 4        b 9 0           b 3 4 9 0 
c 5 6        c 9 0           c 5 6 9 0
d 7 8        e 9 0           
             f 9 0 

我无法产生预期结果的代码是:

My code that does not produce the desired result is:

pd.merge(foo, bar, how='inner', on='x')

相反，代码似乎返回:

 foo           bar            foobar
x y z        x j i           x y z j i
a 1 2        a 9 0           a 1 2 9 0
b 3 4        b 9 0           b 3 4 9 0 
c 5 6        c 9 0           c 5 6 9 0
d 7 8        e 9 0           e * * 9 0
             f 9 0           f * * 9 0

(其中*表示NaN)

我要去哪里错了?我已经到达第三页Google页面，试图解决此问题.无论我做什么，我都会得到一个外部联接，并且两组都包含所有行.

Where am I going wrong? I've already reached the third Google page trying to fix this an nothing works. Whatever I do I get an outer join, with all rows in both sets.

推荐答案

通常，这意味着您在用于连接的列中存在重复项，从而产生了笛卡尔积.

Usually it means that you have duplicates in the column(s) used for joining, resulting in cartesian product.

演示:

In [35]: foo
Out[35]:
   x  y  z
0  a  1  2
1  b  3  4
2  c  5  6
3  d  7  8

In [36]: bar
Out[36]:
   x  j  i
0  a  9  0
1  b  9  0
2  a  9  0
3  a  9  0
4  b  9  0

In [37]: pd.merge(foo, bar)
Out[37]:
   x  y  z  j  i
0  a  1  2  9  0
1  a  1  2  9  0
2  a  1  2  9  0
3  b  3  4  9  0
4  b  3  4  9  0

这篇关于 pandas 内部合并/联接返回所有行的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！