如何比较两个不同长度的Python Pandas系列？

I have two Series of different lengths, and I want to get the indices for which both the indices and the amount are the same in both series.

我有两个不同长度的系列,我想得到两个系列中指数和数量都相同的指数。

Here are the Series:

这是系列:

ipdb> s1
s1
000007720          2000.00
group1            -3732.05
group t3           2432.12
group2           -38147.87
FSHLAJ           -36711.09
EWkayuwo             -3.22
Name: amount, dtype: float64
ipdb> s2
s2
000007720                 2000.00
group1                   -3732.05
group z                  12390.00
group y                  68633.43
group x                     25.00
group w                   3913.00
group v                 -12750.50
group u                    -53.49
group t                  -7500.00
group s                  -1575.82
group r                    -10.00
group q                   1800.00
group p                  -4510.34
EWFhjkaQU                  455.96
group2                  -38147.87
FSHLAJ                  -36711.09
GEKWJ                        5.54
Name: amount, dtype: float64

When I try to compare them, I get:

当我尝试比较它们时,我得到:

ipdb>s1 == s2
*** ValueError: Series lengths must match to compare

How can I achieve my objective?

我怎样才能实现目标?

1 个解决方案

#1

You want to use isin:

你想使用isin:

In [121]:

s2[s2.isin(s1)]
Out[121]:
000007720
group1    -3732.05
group2   -38147.87
FSHLAJ   -36711.09
Name: 2000.00, dtype: float64

I don't know which way round you wanted to perform the comparison, here is the other way:

我不知道你想要进行哪种比较,这是另一种方式:

In [122]:

s1[s1.isin(s2)]
Out[122]:
000007720
group1    -3732.05
group2   -38147.87
FSHLAJ   -36711.09
Name: 2000.00, dtype: float64

The problem with trying to do s1 == s2 is that it doesn't make sense comparing Series or arrays of different lengths.

尝试执行s1 == s2的问题在于比较不同长度的Series或数组是没有意义的。

If you want the indices to match also then add this as a condition:

如果您希望索引也匹配,则将其添加为条件:

In [131]:

s1[(s1.index.isin(s2.index)) & (s1.isin(s2))]
Out[131]:
000007720
group1    -3732.05
group2   -38147.87
FSHLAJ   -36711.09
Name: 2000.00, dtype: float64

#1