w3resource

Pandas: Groupby and aggregate over multiple lists

Pandas Grouping and Aggregating: Split-Apply-Combine Exercise-30 with Solution

Write a Pandas program to split the following dataset using group by on first column and aggregate over multiple lists on second column.

Test Data:

  student_id         marks
0       S001  [88, 89, 90]
1       S001  [78, 81, 60]
2       S002  [84, 83, 91]
3       S002  [84, 88, 91]
4       S003  [90, 89, 92]
5       S003  [88, 59, 90]

Sample Solution:

Python Code :

import pandas as pd
import numpy as np
pd.set_option('display.max_rows', None)
pd.set_option('display.max_columns', None)
df = pd.DataFrame({
    'student_id': ['S001','S001','S002','S002','S003','S003'],
    'marks': [[88,89,90],[78,81,60],[84,83,91],[84,88,91],[90,89,92],[88,59,90]]})
print("Original DataFrame:")
print(df)
print("\nGroupby and aggregate over multiple lists:")
result = df.set_index('student_id')['marks'].groupby('student_id').apply(list).apply(lambda x: np.mean(x,0))
print(result)

Sample Output:

Original DataFrame:
  student_id         marks
0       S001  [88, 89, 90]
1       S001  [78, 81, 60]
2       S002  [84, 83, 91]
3       S002  [84, 88, 91]
4       S003  [90, 89, 92]
5       S003  [88, 59, 90]

Groupby and aggregate over multiple lists:
student_id
S001    [83.0, 85.0, 75.0]
S002    [84.0, 85.5, 91.0]
S003    [89.0, 74.0, 91.0]
Name: marks, dtype: object

Python Code Editor:


Have another way to solve this solution? Contribute your code (and comments) through Disqus.

Previous: Write a Pandas program to split a given dataset using group by on specified column into two labels and ranges.
Next: Write a Pandas program to split the following dataset using group by on ‘salesman_id’ and find the first order date for each group.

What is the difficulty level of this exercise?

Test your Python skills with w3resource's quiz



Python: Tips of the Day

Python: Time library

Time library provides lots of time related functions and methods and is good to know whether you're developing a website or apps and games or working with data science or trading financial markets. Time is essential in most development pursuits and Python's standard time library comes very handy for that.

Let's check out a few simple examples:

moment=time.strftime("%Y-%b-%d__%H_%M_%S",time.localtime())

import time
time_now=time.strftime("%H:%M:%S",time.localtime())
print(time_now)
date_now=time.strftime("%Y-%b-%d",time.localtime())
print(date_now)

Output:

11:36:34
2020-Nov-30