Create len column based on num_type_len by matching with num on two columns.

Expected output:-

| Report_length | Actual_num |
| ------------- | ---------- |
| 75            | 7094       |
| 90            | 1906       |

CodePudding user response：

You can compare splitted values of lists by Actual_num converted to strings and get first match values by next with iter trick:

df['Report_length'] = [next(iter([z.split('_')[-1] 
                       for z in x if z.split('_')[0] == str(y)]), None) 
                       for x, y in zip(df['num_type_len'], df['Actual_num'])]

df = df[['Report_length','Actual_num']]
print (df)
  Report_length  Actual_num
0            75        7094
1            90        1906

Or use DataFrame.explode with lists column with compare splitted values:

df1 = df.explode('num_type_len')

df2 = (df1['num_type_len'].str.split('_', expand=True)
                          .rename(columns={2:'Report_length'})
                          .assign(Actual_num = df1['Actual_num']))

df = df2.loc[df2[0].eq(df2['Actual_num'].astype(str)), ['Report_length', 'Actual_num']]
print (df)
  Report_length  Actual_num
0            75        7094
1            90        1906