I have the following dataframe:

Policy_id Value 
     A           xyz
     B           abc
     A           pqr
     C           lmn

And I want to use np.where() such that whenever the policy_id is equal to A the corresponding value must be appended with a *.

Policy_id    Value 
     A           xyz*
     B           abc
     A           pqr*
     C           lmn

How do I achieve this?

CodePudding user response：

Try

df['new'] = np.where(df['Policy_id'].eq('A'),df['Value'] '*',df['Value'])

CodePudding user response：

If you want to modify your dataframe in place and use np.where:

df['Value']  = np.where(df['Policy_id'].eq('A'), '*', '')

output:

  Policy_id Value
0         A  xyz*
1         B   abc
2         A  pqr*
3         C   lmn

CodePudding user response：

(I just noticed you specifically asked about how to use np.where, but I will leave this Series.mask/Series.where answer for reference.)

• Use Series.mask

  Replace df['Value'] with df['Value'] '*' if df['Policy_id'.eq('A'):

  df['Value'] = df['Value'].mask(df['Policy_id'].eq('A'), df['Value']   '*')
  

• Or Series.where

  Inverse of masking, so flip the condition from eq to ne:

  df['Value'] = df['Value'].where(df['Policy_id'].ne('A'), df['Value']   '*')
  

Note that inplace is possible but not recommended and will soon be deprecated:

df['Value'].mask(df['Policy_id'].eq('A'), df['Value']   '*', inplace=True)