Using multi-metric filtering to cross-promote books
As a final request, the founder of the ebook selling start-up asks you to perform additional filtering. Your previous attempt returned 82 rules, but she wanted only one. The rules
dataset has again been made available in the console. Finally, Zhang's metric has been computed for you and included in the rules
DataFrame under the column header zhang
.
This exercise is part of the course
Market Basket Analysis in Python
Exercise instructions
- Set the lift threshold to be greater than 1.5.
- Use a conviction threshold of 1.0.
- Require Zhang's metric to be greater than 0.65.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Set the lift threshold to 1.5
rules = rules[rules['____'] > ____]
# Set the conviction threshold to 1.0
rules = rules[____]
# Set the threshold for Zhang's rule to 0.65
rules = ____
# Print rule
print(rules[['antecedents','consequents']])