pandas.Series.drop

Series.drop 方法可以返回一个新对象，移除指定的 index labels.

import pandas as pd
import numpy as nps = pd.Series(np.arange(5.), index=['a', 'b', 'c', 'd', 'e'])
s
"""
a    0.0
b    1.0
c    2.0
d    3.0
e    4.0
dtype: float64
"""

s2 = s.drop('c')
s2
"""
a    0.0
b    1.0
d    3.0
e    4.0
dtype: float64
"""

s.drop(['d', 'c'])
"""
a    0.0
b    1.0
e    4.0
dtype: float64
"""

pandas.DataFrame.drop

对于 DataFrame，我们可以通过指定 axis 参数来决定从哪个轴进行 drop。

df = pd.DataFrame(np.arange(16).reshape((4, 4)),index=['BeiJ', 'ShangH', 'ShenZ', 'GuangZ'],columns=['one', 'two', 'three', 'four'])
print(df)
"""one  two  three  four
BeiJ      0    1      2     3
ShangH    4    5      6     7
ShenZ     8    9     10    11
GuangZ   12   13     14    15
"""

只传入一个序列，则丢弃的是 index：

df.drop(['BeiJ', 'ShenZ'])
"""one  two  three  four
ShangH    4    5      6     7
GuangZ   12   13     14    15
"""

df.drop(['BeiJ', 'ShenZ']) 其实是 df.drop(index=['BeiJ', 'ShenZ'])

将 axis 参数设为 1 或者 ‘columns’，则会丢弃 column：

df.drop(['one', 'three'], axis=1)
"""two  four
BeiJ      1     3
ShangH    5     7
ShenZ     9    11
GuangZ   13    15
"""

以上语句和下面的表达是等价的：

df.drop(columns=['one', 'three'])
"""two  four
BeiJ      1     3
ShangH    5     7
ShenZ     9    11
GuangZ   13    15
"""

如果我们不想要创建新的对象，而是直接在原对象基础之上修改，则可以指定 inplace=True：

df.drop(['BeiJ', 'ShenZ'], inplace=True)
print(df)
"""one  two  three  four
ShangH    4    5      6     7
GuangZ   12   13     14    15
"""

References

[1] NumPy Reference. https://numpy.org/doc/stable/reference/index.html
[2] Python for Data Analysis, 2nd^{\rm nd}nd edition. Wes McKinney.