Pandas: Conditionally replace values based on other columns values

Question

I have a dataframe (df) that looks like this:

                    environment     event   
time                    
2017-04-28 13:08:22     NaN         add_rd  
         


        

Answer 1

if you want to replace just 'add_rd' with 'RD', this can be useful to you

keys_to_replace = {'add_rd':'RD','add_env':'simple'}
df['environment'] = df.groupby(['event'])['environment'].fillna(keys_to_replace['add_rd'])
df

output:

    environment event
0   RD          add_rd
1   RD          add_rd
2   test        add_env
3   prod        add_env

if you have many values to replace based on event, then you may need to follow groupby with 'event' column values

keys_to_replace = {'add_rd':'RD','add_env':'simple'}
temp = df.groupby(['event']).apply(lambda x:  x['environment'].fillna(keys_to_replace[x['event'].values[0]]))
temp.index = temp.index.droplevel(0)
df['environment'] = temp.sort_index().values

output:

   environment  event
0   RD          add_rd
1   RD          add_rd
2   test        add_env
3   prod        add_env