CODEWITHSUNDEEP
pythonpandasconcatenation
n8-da-gr8
n8-da-gr8

Reputation: 551

Pandas - conditionally concat two columns

Given a dataframe

Patient ID     Instructions    ID Replaced
   1                N/A           ID123
   2                              ID124
   3                              ID125
   4                xyz           ID126
   5                xyz           ID127
   6                              ID128
   7                Replacement   ID129
   8                Replace       ID130
   9                replaced      ID131
   10               xyz           ID132

How can I make a new column, which concats Instructions with ID Replaced if replac substring is found?

Patient ID  Instructions    ID Replaced     Comments
    1           N/A            ID123    
    2                          ID124    
    3                          ID125    
    4           xyz            ID126    
    5           xyz            ID127    
    6                          ID128    
    7           Replacement    ID129    Replacement | ID129
    8           Replace        ID130    Replace | ID130
    9           Replaced       ID131    Replaced | ID131
    10          xyz            ID132

I've tried the following but the Comments column is completely empty

mani_df['Comments'] = ""
# if instructions contains 'replac' , concat with ID replaced 
if "replace" in df['Instructions']:
    df['Comments'] = df['Instructions'].str.cat(df['ID Replaced'], sep = " | ")

and I tried to use a boolean mask but this returns False for the first two rows

mask = mani_df['Special Handling Directions'].str.contains('replac')

    Out[55]: 
    0    False
    1    False
    2      NaN
    3      NaN

Upvotes: 0

Views: 3505

Answers (1)

rafaelc
rafaelc

Reputation: 59304

You may usestr.contains with case=False and just concat using pandas indexing

mask = df.Instructions.str.contains('Replace', case=False).fillna(False)

df['Comments'] = df.loc[mask, 'Instructions'] + ' | ' + df['ID Replaced']

Of course, you can fillna at the end to get empty strings (which looks like your expected output)

df.fillna('')

Yields

    Patient ID  Instructions    ID Replaced Comments
0   1                           ID123   
1   2                           ID124       
2   3                           ID125       
3   4           xyz             ID126   
4   5           xyz             ID127   
5   6                           ID128       
6   7           Replacement     ID129       Replacement | ID129
7   8           Replace         ID130       Replace | ID130
8   9           replaced        ID131       replaced | ID131
9   10          xyz             ID132

Upvotes: 2

Related Questions