python-3.x - 如何将列的内容移动到它旁边的列

Question

数据框如下所示:

import pandas as pd 

df = pd.DataFrame ({
    'Name':['Brian','John','Adam'],
    'HomeAddr':[12,32,44],
    'Age':['M','M','F'],
    'Genre': ['NaN','NaN','NaN']
    })

当前输出为:

    Name  HomeAddr Age Genre
0  Brian        12   M   NaN
1   John        32   M   NaN
2   Adam        44   F   NaN

我想以某种方式将 HomeAddr 和 Age 列的内容转移到 column+1。以下是预期输出的示例。

    Name HomeAddr  Age Genre
0  Brian      NaN   12     M
1   John      NaN   32     M
2   Adam      NaN   44     F

我尝试使用 .shift() 但它不起作用。

import pandas as pd 

df = pd.DataFrame ({
    'Name':['Brian','John','Adam'],
    'HomeAddr':[12,32,44],
    'Age':['M','M','F'],
    'Genre': ['NaN','NaN','NaN']
    })

df['HomeAddr'] = df['HomeAddr'].shift(-1)
print(df)


    Name  HomeAddr Age Genre
0  Brian      32.0   M   NaN
1   John      44.0   M   NaN
2   Adam       NaN   F   NaN

有什么想法吗？谢谢!

Answer 1

使用 DataFrame.shift ，但有必要将列转换为字符串以避免丢失值，然后将数字列转换回:

df.loc[:, 'HomeAddr':] = df.loc[:, 'HomeAddr':].astype(str).shift(1, axis=1)
df['Age'] = pd.to_numeric(df['Age'])
print (df)
    Name HomeAddr Age Genre
0  Brian      NaN  12     M
1   John      NaN  32     M
2   Adam      NaN  44     F

另一个开箱即用的解决方案:

df = df.drop('Genre', axis=1).rename(columns={'HomeAddr':'Age', 'Age':'Genre'})
df.insert(1, 'HomeAddr', np.nan)
print (df)
    Name  HomeAddr  Age Genre
0  Brian       NaN   12     M
1   John       NaN   32     M
2   Adam       NaN   44     F