在df-python中，将列值转换为两个不同的列

Question

我有一个包含4列的df：ID、Value、UpperBound和LowerBound，如下所示：

ID | Value | UpperBound | LowerBound
1     23        NaN         NaN
2     55        NaN         NaN
3     87        NaN         NaN
4     99        NaN         NaN
5     NaN       50          5
6     NaN       5           1
7     NaN       95          50
8     NaN       99          95

我想根据各自的边界将Value列中的值转换为UpperBound和LowerBound(例如，23适合50-5(。所以输出看起来像这样：

ID | Value | UpperBound | LowerBound
1     NaN        50         5
2     NaN        95         50
3     NaN        95         50
4     NaN        99         95
5     NaN       50          5
6     NaN       5           1
7     NaN       95          50
8     NaN       99          95

Value列最终将只包含NaN值，因此我稍后可以消除它。UpperBound和LowerBound如下：

1-5
5-50
50-95
95-99

有人能帮我创建必要的代码吗？

提前非常感谢！！

Answer 1

假设原始数据帧如下所示：

import pandas as pd
import numpy as np
df = pd.DataFrame({'Value': [23,55,87,99,999],
'UpperBound': [np.nan, np.nan, np.nan, np.nan, np.nan],
'LowerBound': [np.nan, np.nan, np.nan, np.nan, np.nan] })
print(df)
Value  UpperBound  LowerBound
0     23         NaN         NaN
1     55         NaN         NaN
2     87         NaN         NaN
3     99         NaN         NaN
4    999         NaN         NaN

这里有一种方法，假设包含下限，排除上限：

def check_bound(x):
for (a,b) in [(1,5), (5,50), (50,95), (95,100)]:
if x in np.arange(a,b):
return [a,b]      #this will exit the function
return [np.nan, np.nan]   #iteration ended means bound is not found
df[['LowerBound', 'UpperBound']] = df['Value'].apply(lambda x: check_bound(x)).tolist()
print(df)
Value  UpperBound  LowerBound
0     23        50.0         5.0
1     55        95.0        50.0
2     87        95.0        50.0
3     99       100.0        95.0
4    999         NaN         NaN