6.4 Grouping and Summarising | Code Scholars Tutoring

How it works

Grouping splits rows by a key, applies an aggregation to each group, and combines the results into a new table.

Use grouping to answer per-category questions rather than looping over categories manually.

Common error

Reporting an average over groups of very different sizes without saying so, which hides the imbalance.

Worked example

These examples require NumPy or pandas, as shown by the imports.

Python

import pandas as pd
df = pd.DataFrame({"team": ["A", "B", "A"], "score": [4, 9, 8]})
means = df.groupby("team")["score"].mean()
for team in sorted(means.index):
    print(team, float(means[team]))

Output

A 6.0
B 9.0

Grouping computes separate summaries rather than mixing categories. Include group sizes when they affect how confidently the results can be interpreted.

Exercises

6.4.1 · Explain

Why can averaging group averages give the wrong overall mean when group sizes differ?

Show solution

It gives equal weight to groups rather than observations. Use counts and sums, or an appropriate weighted mean, when calculating the overall observation mean.

6.4.2 · Write code

For groups A, A, B with values 2, 6, 5, print each group total in sorted order.

Show solution

Group by the category and sum the value column.

One valid solution

import pandas as pd
df = pd.DataFrame({"group": ["A", "A", "B"], "value": [2, 6, 5]})
totals = df.groupby("group")["value"].sum()
for group in sorted(totals.index):
    print(group, int(totals[group]))

Expected output

A 8
B 5

Open Python compiler