How to concat Pandas dataframe columns

Question

How to concat columns of different data types in a Pandas dataframe such that if column number is concatenated with column operator, and I do a groupby('user').sum(), I can have the appropriate aggregation:

    number   operator user
id                      
1      193        -    A
2      332        +    B
3     4434        +    A
4      432        -    C
5      652        +    C
6      567        +    D


#after concat:

    number operator user
id                      
1     -193        -    A
2      332        +    B
3     4434        +    A
4     -432        -    C
5      652        +    C
6      567        +    D


#df.groupby('user').sum()

      number
user        
A       4241
B        332
C        220
D        567

Answer 1

Use loc with boolean mask to make the 'number' values negative:

In [34]:
df.loc[df['operator'] == '-', 'number'] = -df['number']
df

Out[34]:
    number operator user
id                      
1     -193        -    A
2      332        +    B
3     4434        +    A
4     -432        -    C
5      652        +    C
6      567        +    D

You can then groupby on 'user' can call sum on 'number' column:

In [35]:    
df.groupby('user')['number'].sum()

Out[35]:
user
A    4241
B     332
C     220
D     567
Name: number, dtype: int64