Group and multiply columns with conditions
Question:
I’m trying to multiply 2 columns until get a desired value(8), but i need to group first, also need to keep the first mult if the values is already under the desired valued (This part is the problematic)
MPRO
ID
Nuevo_I
Nuevo_P
1
ID1
5
3
1
ID1
2
3
1
ID1
1
3
1
ID2
2
3
1
ID2
1
3
1
ID3
3
2
2
ID1
3
4
2
ID2
3
2
2
ID2
3
1
2
ID2
2
1
Desired Output:
Macroproceso
ID
Nuevo_I
Nuevo_P
F
1
ID1
5
3
15
1
ID1
2
3
6
1
ID1
1
3
DONE
1
ID2
2
3
6
1
ID2
1
3
DONE
1
ID3
3
2
6
2
ID1
3
4
12
2
ID2
3
2
6
2
ID2
3
1
DONE
2
ID2
2
1
DONE
I have tried with this code:
A = (
df['Nuevo_I'].mul(df['Nuevo_P'])
.groupby([df['MPRO'], df['ID']])
)
B = A.le(8)..groupby([df['MPRO'], df['ID']]).shift(fill_value=False)
Hope anyone could give me hand, thanks!
Answers:
You can use a groupby.apply:
# compute the product
s = df[['Nuevo_I', 'Nuevo_P']].prod(axis=1)
# identify values after the first value < 8
m = (s.le(8).groupby([df['MPRO'], df['ID']], group_keys=False)
.apply(lambda x: x.shift(fill_value=False).cummax())
)
# mask
df['F'] = s.mask(m, 'DONE')
Output:
MPRO ID Nuevo_I Nuevo_P F
0 1 ID1 5 3 15
1 1 ID1 2 3 6
2 1 ID1 1 3 DONE
3 1 ID2 2 3 6
4 1 ID2 1 3 DONE
5 1 ID3 3 2 6
6 2 ID1 3 4 12
7 2 ID2 3 2 6
8 2 ID2 3 1 DONE
9 2 ID3 2 1 2
I’m trying to multiply 2 columns until get a desired value(8), but i need to group first, also need to keep the first mult if the values is already under the desired valued (This part is the problematic)
| MPRO | ID | Nuevo_I | Nuevo_P |
|---|---|---|---|
| 1 | ID1 | 5 | 3 |
| 1 | ID1 | 2 | 3 |
| 1 | ID1 | 1 | 3 |
| 1 | ID2 | 2 | 3 |
| 1 | ID2 | 1 | 3 |
| 1 | ID3 | 3 | 2 |
| 2 | ID1 | 3 | 4 |
| 2 | ID2 | 3 | 2 |
| 2 | ID2 | 3 | 1 |
| 2 | ID2 | 2 | 1 |
Desired Output:
| Macroproceso | ID | Nuevo_I | Nuevo_P | F |
|---|---|---|---|---|
| 1 | ID1 | 5 | 3 | 15 |
| 1 | ID1 | 2 | 3 | 6 |
| 1 | ID1 | 1 | 3 | DONE |
| 1 | ID2 | 2 | 3 | 6 |
| 1 | ID2 | 1 | 3 | DONE |
| 1 | ID3 | 3 | 2 | 6 |
| 2 | ID1 | 3 | 4 | 12 |
| 2 | ID2 | 3 | 2 | 6 |
| 2 | ID2 | 3 | 1 | DONE |
| 2 | ID2 | 2 | 1 | DONE |
I have tried with this code:
A = (
df['Nuevo_I'].mul(df['Nuevo_P'])
.groupby([df['MPRO'], df['ID']])
)
B = A.le(8)..groupby([df['MPRO'], df['ID']]).shift(fill_value=False)
Hope anyone could give me hand, thanks!
You can use a groupby.apply:
# compute the product
s = df[['Nuevo_I', 'Nuevo_P']].prod(axis=1)
# identify values after the first value < 8
m = (s.le(8).groupby([df['MPRO'], df['ID']], group_keys=False)
.apply(lambda x: x.shift(fill_value=False).cummax())
)
# mask
df['F'] = s.mask(m, 'DONE')
Output:
MPRO ID Nuevo_I Nuevo_P F
0 1 ID1 5 3 15
1 1 ID1 2 3 6
2 1 ID1 1 3 DONE
3 1 ID2 2 3 6
4 1 ID2 1 3 DONE
5 1 ID3 3 2 6
6 2 ID1 3 4 12
7 2 ID2 3 2 6
8 2 ID2 3 1 DONE
9 2 ID3 2 1 2