train-model.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
Script to train and save the book recommendation model.
"""

from utils import train_and_save_model
import pandas as pd

if __name__ == "__main__":
    # Check the dataset for duplicates first
    df = pd.read_csv('books_summary.csv')
    duplicates = df.duplicated().sum()
    print(f"Number of duplicate rows in dataset: {duplicates}")
    
    # Check for books with the same title
    title_counts = df['book_name'].value_counts()
    duplicate_titles = title_counts[title_counts > 1]
    if not duplicate_titles.empty:
        print(f"Found {len(duplicate_titles)} book titles that appear multiple times.")
        print("Examples:")
        for title, count in duplicate_titles.head(5).items():
            print(f"- '{title}' appears {count} times")
    else:
        print("No duplicate book titles found.")
    
    # Train and save the model
    result = train_and_save_model('books_summary.csv')
    
    if "error" in result:
        print(f"Error: {result['error']}")
    else:
        print("Model training completed successfully!")