Why clustering groups similar data in SciPy - See It in Action

Why clustering groups similar data

📖 Scenario: Imagine you have a small shop and you want to group your customers based on their shopping habits. Grouping similar customers helps you understand their preferences better and offer them personalized deals.

🎯 Goal: You will create a simple dataset of customer shopping amounts, set a threshold for grouping, apply clustering using scipy, and then display the groups formed.

📋 What You'll Learn

Create a list of customer shopping amounts

Set a distance threshold for clustering

Use scipy.cluster.hierarchy to cluster the data

Print the cluster labels for each customer

💡 Why This Matters

🌍 Real World

Clustering helps businesses group customers with similar behaviors to target marketing and improve service.

💼 Career

Data scientists use clustering to find patterns in data, segment customers, and make data-driven decisions.

1

Create the customer data list

Create a list called shopping_amounts with these exact values: 5, 7, 15, 18, 25, 30.

SciPy

# Create the list shopping_amounts with the given values
# Your code here

Need a hint?

Use square brackets to create a list and separate numbers with commas.

2

Set the clustering distance threshold

Create a variable called distance_threshold and set it to 10.

SciPy

shopping_amounts = [5, 7, 15, 18, 25, 30]
# Set distance_threshold to 10
# Your code here

Need a hint?

Just assign the number 10 to the variable distance_threshold.

3

Cluster the shopping amounts

Import linkage and fcluster from scipy.cluster.hierarchy. Use linkage on shopping_amounts with method 'complete' and store in linked. Then create clusters using fcluster with linked, distance_threshold, and criterion 'distance'.

SciPy

shopping_amounts = [5, 7, 15, 18, 25, 30]
distance_threshold = 10
# Import linkage and fcluster from scipy.cluster.hierarchy
# Create linked using linkage on shopping_amounts with method 'complete'
# Create clusters using fcluster with linked, distance_threshold, and criterion 'distance'
# Your code here

Need a hint?

Convert each number to a list like [x] before clustering because linkage expects 2D data.

4

Print the cluster labels

Print the clusters list to show which group each shopping amount belongs to.

SciPy

from scipy.cluster.hierarchy import linkage, fcluster

shopping_amounts = [5, 7, 15, 18, 25, 30]
distance_threshold = 10

linked = linkage([[x] for x in shopping_amounts], method='complete')
clusters = fcluster(linked, distance_threshold, criterion='distance')
# Print the clusters list
# Your code here

Need a hint?

Use print(clusters.tolist()) to display the cluster labels as a list.