pychallenger

pandas II

GroupBy and Aggregation

Learn how to use groupby() to split a DataFrame into groups and apply aggregation functions like size(), mean(), and max().

groupby() splits a DataFrame into groups based on a column's values, then lets you apply an aggregation function to each group.

Count how many companies exist per industry using groupby + size():

import pandas as pd
 
df = pd.read_csv('/data/companies.csv')
 
counts = df.groupby(
    'Industry'
).size()
print(counts.head(5))

Python

Output

What will be the output?

Python

 The number of rows per industry
 Output

 The average revenue per industry
 Output

 A filtered DataFrame for each industry
 Output

Chain an aggregation method like mean(), sum(), or max() to compute a statistic per group:

import pandas as pd
 
df = pd.read_csv('/data/companies.csv')
 
avg_rev = df.groupby(
    'Industry'
)['Revenue (USD)'].mean()
print(avg_rev.head(4))

Python

Output

What will be the output?

Python

 The mean revenue for each industry
 Output

 All revenue values sorted by industry
 Output

 The industry with the highest mean revenue
 Output

Sort the grouped result to find the top industry by revenue:

import pandas as pd
 
df = pd.read_csv('/data/companies.csv')
 
top = df.groupby(
    'Industry'
)['Revenue (USD)'].mean()
.sort_values(ascending=False)
print(top.head(3))

Python

Output

What will be the output?

Python

 size() or count()
 Output

 len()
 Output

 sum()
 Output

What will be the output?

Python

 The maximum value of val for each group in col
 Output

 The row with the maximum value overall
 Output

 The number of groups
 Output

What will be the output?

Python

 Yes — the result is a Series that can be sorted
 Output

 No — groupby results cannot be sorted
 Output

 Only with reset_index() first
 Output

Imprint Privacy Policy Terms