Home / Questions / 552471
Accepted
efueyo
Asked: 2022-08-19 08:21:00 +0800 CST 2022-08-19 08:21:00 +0800 CST

Python groupby().sum(). How to apply it to only some selected columns?

  • 772

I have the following DataFrame.

import pandas as pd

lista = [['Producto1', '06/22/2012', 66.72721, 17.995],
 ['Producto1', '09/18/2017', 240.56891, 19.244],
 ['Producto1', '01/08/2018', 219.24285, 17.459],
 ['Producto1', '03/06/2018', 667.32977, 18.245],
 ['Producto2', '10/07/2018', 49.328, 254.64],
 ['Producto3', '27/04/2016', 733.27266, 13.7643],
 ['Producto3', '06/12/2019', 1213.05103, 13.71],
 ['Producto4', '10/08/2012', 44986.57615, 10.4718] ]

df_aux = pd.DataFrame(lista)
df_aux.columns = ["Producto", "Fecha", "Cantidad", 'Precio']
df_aux['Total'] = df_aux['Cantidad'] * df_aux['Precio']
df_aux

Bring back.

    Producto    Fecha   Cantidad    Precio  Total
0   Producto1   06/22/2012  66.72721    17.9950     1200.756144
1   Producto1   09/18/2017  240.56891   19.2440     4629.508104
2   Producto1   01/08/2018  219.24285   17.4590     3827.760918
3   Producto1   03/06/2018  667.32977   18.2450     12175.431654
4   Producto2   10/07/2018  49.32800    254.6400    12560.881920
5   Producto3   27/04/2016  733.27266   13.7643     10092.984874
6   Producto3   06/12/2019  1213.05103  13.7100     16630.929621
7   Producto4   10/08/2012  44986.57615     10.4718     471090.4281

The DataFrame has four columns: 'Products', 'Dates', 'Quantity', 'Price', 'Total' I want to group by products, and obtain a DataFrame with only two columns, the sum of 'Quantity' and 'Total. I execute

df_productos = df_aux.groupby(by= ['Producto']).sum( )
df_productos

Which brings me back

    Cantidad    Precio  Total
Producto            
Producto1   1193.86874  72.9430     21833.456820
Producto2   49.32800    254.6400    12560.881920
Producto3   1946.32369  27.4743     26723.914495
Producto4   44986.57615     10.4718     471090.428128

With the groupiby() method, can I get it not to return (add) the prices column? I will appreciate suggestions.

python-3.x

2 Answers

  1. Best Answer

    To define which columns and aggregations you want to use you can use the aggpandas groups method.

    df_productos = df_aux.groupby(by=['Producto']).agg({"Cantidad": "sum", "Total": "sum"})
df_productos
    • 3

  2. You can use the ilocand/or function locto select the columns you need. Something like:

    df_productos.loc[:, ["Producto", "Cantidad"]]
    • 2