Question

如何根据名称中的某个字符串删除数据框列？

示例：

           house1    house2    chair1  chair2
index
  1         foo       lee       sam      han
  2         fowler    smith     had      sid
  3         cle       meg       mag      mog

我想删除字符串中包含'chair'的列。如何以有效的方式完成？感谢。

Answer 1

auto-complete

Answer 2

<强> UPDATE2：

In [315]: df
Out[315]:
   3M110%  3M80% 6M90% 6M95% 1N90% 2M110% 3M95%
1     foo    lee   sam   han   aaa    aaa   fff
2  fowler  smith   had   sid   aaa    aaa   fff
3     cle    meg   mag   mog   aaa    aaa   fff

In [316]: df.loc[:, ~df.columns.str.contains('90|110')]
Out[316]:
   3M80% 6M95% 3M95%
1    lee   han   fff
2  smith   sid   fff
3    meg   mog   fff

<强>更新

In [40]: df
Out[40]:
   house1 house2 chair1 chair2 door1 window1 floor1
1     foo    lee    sam    han   aaa     aaa    fff
2  fowler  smith    had    sid   aaa     aaa    fff
3     cle    meg    mag    mog   aaa     aaa    fff

In [41]: df.filter(regex='^(?!(chair|door|window).*?)')
Out[41]:
   house1 house2 floor1
1     foo    lee    fff
2  fowler  smith    fff
3     cle    meg    fff

原始回答：

这里有一些选择：

In [37]: df.drop(df.filter(like='chair').columns, 1)
Out[37]:
   house1 house2
1     foo    lee
2  fowler  smith
3     cle    meg

In [38]: df.filter(regex='^(?!chair.*)')
Out[38]:
   house1 house2
1     foo    lee
2  fowler  smith
3     cle    meg

Answer 3

这应该这样做：

df.drop([col for col in df.columns if 'chair' in col],axis=1,inplace=True)

时序

MaxU方法2

Answer 4

还有一个选择：

import pandas as pd

df = pd.DataFrame({'house1':['foo','fowler','cle'],
                   'house2':['lee','smith','meg'],
                   'chair1':['sam','had','mag'],
                   'chair2':['han','sid','mog']})

mask = ['chair' not in x for x in df]

df = df[df.columns[mask]]

根据字符串条件删除列

4 个答案:

时序