The Most Common Value

Part of the Statistics for ML section of Coddy's Machine Learning journey. Lesson 5 of 64.

T-shirt sizes are labels: there is no mean of M and XL. A category column's center is its mode, the most common value:

Bars counting 8 T-shirt orders by size: S 2, M 3, L 2, XL 1. The tallest bar, M, is dark and labelled mode

.mode() finds it:

sizes = pd.Series(["M", "L", "M", "S", "XL", "M", "L", "S"])
print(sizes.mode()[0])

This prints:

Values can tie, so .mode() returns a Series of every winner, sorted. [0] takes the first.

The mode fills gaps in a category column. It is also the simplest model of all: always guess the most common class.

Challenge

Easy

A ski shop wants to know what to keep in stock. rentals.csv logs 30 rentals with the columns item, boot_size, level and days.

For each column, in order, print its name, its mode, and how many rentals have that value.
Then compare the mean of days, rounded to 1 decimal, with its mode.

For a shorter log of 8 rentals:

item: snowboard (5 of 8)
boot_size: 42 (4 of 8)
level: beginner (5 of 8)
days: 1 (4 of 8)
Days rented, mean vs mode: 2.8 vs 1

No rental lasts 2.8 days: the mode is a value that really happens.

Try it yourself

import pandas as pd

rentals = pd.read_csv("rentals.csv")

# Write your code below

Test yourself

This lesson includes a short quiz. Start the lesson to answer it and track your progress.

All lessons in Statistics for ML

1Center and Spread

4Samples and Populations

7Trend Lines

10Encoding Categories

3Probability by Simulation

6Relationships

9Feature Scaling

12Final Challenges