tempCodeRunnerFile.python

import pandas as pd
from mlxtend.frequent_patterns import apriori, association_rules

# Hardcoded fitness products (replace with your actual items)
fitness_products = {
    "transaction_1": ["Protein Powder", "Yoga Mat", "Water Bottle"],
    "transaction_2": ["Running Shoes", "Jump Rope", "Fitness Tracker"],
    "transaction_3": ["Protein Powder", "Gym Bag", "Resistance Bands"],
    "transaction_4": ["Running Shoes", "Yoga Mat", "Water Bottle"],
    "transaction_5": ["Jump Rope", "Gym Bag", "Fitness Tracker"],
}

# Convert transactions to a list of itemsets
transactions = list(fitness_products.values())

# Define minimum support count (adjust as needed)
min_support = 0.3  # Consider adjusting based on number of transactions

# Apply Apriori algorithm
itemsets_frequent = apriori(transactions, min_support=min_support)

# Get the frequent itemsets with support counts
frequent_itemsets_with_support = itemsets_frequent[["itemsets", "support"]]

# Display frequent itemsets and their support counts
print("Frequent Itemsets (with support):")
print(frequent_itemsets_with_support.to_string(index=False))

# Optional: Generate association rules with minimum confidence
min_confidence = 0.5  # Adjust as needed

if len(frequent_itemsets_with_support) > 1:  # Check if enough itemsets for rules
    rules = association_rules(
        frequent_itemsets_with_support, metric="confidence", min_threshold=min_confidence
    )

    # Sort rules by confidence in descending order
    rules = rules.sort_values(by=["confidence"], ascending=False)

    # Display association rules (if generated)
    print("\nAssociation Rules (with confidence):")
    print(rules[["antecedents", "consequents", "confidence"]].to_string(index=False))
else:
    print("\nNot enough frequent itemsets to generate association rules.")