Replace a specific value with another using Pandas-CodePudding

I would like to replace a specific value or values with another value.

Data

ID  Date    hi     hello
AA  Q4.2022 1      0
BB  Q4.2022 1      1
CC  Q4.2022 HI111  1

Desired

ID  Date    hi     hello
AA  Q4.2022 1      0
BB  Q4.2022 1      1
CC  Q4.2022        1

Doing

I have using this statement, however, this deletes the all the values

df['hi'] = df['hi'].str.replace('HI111','')

Any suggestion is appreciated.

CodePudding user response：

Perhaps, the numbers are int types; then you could try to_numeric isna and use it in mask:

df['hi'] = df['hi'].mask(pd.to_numeric(df['hi'], errors='coerce').isna(), '')

or if you want to change the type of the numbers to strings as well, you could use to_numeric fillna astype(str):

df['hi'] = (pd.to_numeric(df['hi'], errors='coerce').fillna('')
            .astype(str).str.split('.').str[0])

Output:

   ID     Date hi  hello
0  AA  Q4.2022  1      0
1  BB  Q4.2022  1      1
2  CC  Q4.2022         1