You can use the following syntax to find the sum of rows in a pandas DataFrame that meet some criteria:
#find sum of each column, grouped by one column
df.groupby('group_column').sum()
#find sum of one specific column, grouped by one column
df.groupby('group_column')['sum_column'].sum()
The following examples show how to use this syntax with the following data frame:
import pandas as pd
#create DataFrame
df = pd.DataFrame({'team': ['a', 'a', 'b', 'b', 'b', 'c', 'c'],
'points': [5, 8, 14, 18, 5, 7, 7],
'assists': [8, 8, 9, 3, 8, 7, 4],
'rebounds': [1, 2, 2, 1, 0, 4, 1]})
#view DataFrame
df
team points assists rebounds
0 a 5 8 1
1 a 8 8 2
2 b 14 9 2
3 b 18 3 1
4 b 5 8 0
5 c 7 7 4
6 c 7 4 1
Example 1: Perform a SUMIF Function on One Column
The following code shows how to find the sum of points for each team:
df.groupby('team')['points'].sum()
team
a 13
b 37
c 14
This tells us:
- Team ‘a’ scored a total of 13 points
- Team ‘b’ scored a total of 37 points
- Team ‘c’ scored a total of 14 points
Example 2: Perform a SUMIF Function on Multiple Columns
The following code shows how to find the sum of points and rebounds for each team:
df.groupby('team')[['points', 'rebounds']].sum()
points rebounds
team
a 13 3
b 37 3
c 14 5
Example 3: Perform a SUMIF Function on All Columns
The following code shows how to find the sum of all columns in the data frame for each team:
df.groupby('team').sum()
points assists rebounds
team
a 13 16 3
b 37 20 3
c 14 11 5
Additional Resources
How to Perform a COUNTIF Function in Pandas
How to Count Observations by Group in Pandas
How to Find the Max Value by Group in Pandas