[英]Merging values of two columns of dataframe with different dtypes
I have the following pandas dataframe
: 我有以下pandas dataframe
:
import pandas as pd
df = pd.DataFrame({"pos": [1, 2, 3], "chain": ["A", "B", "C"]})
Giving: 赠送:
chain pos
0 A 1
1 B 2
2 C 3
and df.types
: 和df.types
:
chain object
pos int64
dtype: object
I'm looking for a way to merge Series df["chain"]
and df["pos"]
to have the following: 我正在寻找一种合并系列df["chain"]
和df["pos"]
来实现以下目标:
chain+pos
0 A1
1 B2
2 C3
and df.dtypes
: 和df.dtypes
:
chain+pos object
dtype: object
Is there an easy way to do it? 有一个简单的方法吗?
df.astype(str).sum(1)
Out[489]:
0 A1
1 B2
2 C3
dtype: object
In [34]: df['chain'] += df.pop('pos').astype(str)
In [35]: df
Out[35]:
chain
0 A1
1 B2
2 C3
renaming column: 重命名列:
In [37]: df = df.rename(columns={'chain':'chain+pos'})
In [38]: df
Out[38]:
chain+pos
0 A1
1 B2
2 C3
The solution by MaxU works very well. MaxU的解决方案非常有效。 Otherwise you can use the following also 否则您也可以使用以下内容
df["chain+pos"] = df['chain'] + df['pos'].map(str)
After this, you have to drop df['chain'] and df['pos'] to attain the desired result. 在此之后,您必须删除df ['chain']和df ['pos']以获得所需的结果。
----------------- Edit -----------------编辑
As @MaxU pointed out in his comment below, here is a concise way of achieving the desired result - 正如@MaxU在下面的评论中所指出的,这是一种实现理想结果的简洁方法 -
df['chain+pos'] = df.pop('chain') + df.pop('pos').map(str)
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.