demo.py

"""
Demo script for Book Popularity Predictor
==========================================

This script demonstrates the machine learning model capabilities
without the web interface. Perfect for showcasing the core functionality.
"""

import sys
import os
sys.path.append(os.path.join(os.path.dirname(__file__), 'src'))

from src.prediction_utils import BookPopularityPredictor
import pandas as pd

def run_demo():
    """Run demonstration of the book popularity prediction model"""
    
    print("📚 Book Popularity Predictor - Demo")
    print("=" * 50)
    
    # Initialize predictor
    predictor = BookPopularityPredictor(models_dir='models')
    
    # Load model
    if not predictor.load_model_components():
        print("❌ Failed to load model. Please ensure the model is trained.")
        print("Run: python src/model_training.py")
        return
    
    print("✅ Model loaded successfully!")
    print(f"📊 Available authors: {len(predictor.get_top_authors())}")
    
    # Demo predictions
    demo_cases = [
        {
            "author": "Stephen King",
            "ratings_count": 100000,
            "reviews_count": 12000,
            "description": "Popular Stephen King horror novel"
        },
        {
            "author": "J.K. Rowling", 
            "ratings_count": 500000,
            "reviews_count": 60000,
            "description": "New Harry Potter series book"
        },
        {
            "author": "Jane Austen",
            "ratings_count": 75000,
            "reviews_count": 9000,
            "description": "Classic romance novel reprint"
        },
        {
            "author": "Rick Riordan",
            "ratings_count": 200000,
            "reviews_count": 25000,
            "description": "New Percy Jackson adventure"
        }
    ]
    
    print("\\n🎯 Demo Predictions:")
    print("-" * 50)
    
    for i, case in enumerate(demo_cases, 1):
        print(f"\\n{i}. {case['description']}")
        print(f"   Author: {case['author']}")
        print(f"   Expected Ratings: {case['ratings_count']:,}")
        print(f"   Expected Reviews: {case['reviews_count']:,}")
        
        # Make prediction
        result = predictor.predict_book_rating(
            case['author'], 
            case['ratings_count'], 
            case['reviews_count']
        )
        
        if "error" in result:
            print(f"   ❌ Error: {result['error']}")
        else:
            predicted_rating = result["predicted_rating"]
            confidence = result["confidence"]
            
            # Rating interpretation
            if predicted_rating >= 4.5:
                rating_desc = "Excellent (⭐⭐⭐⭐⭐)"
            elif predicted_rating >= 4.0:
                rating_desc = "Very Good (⭐⭐⭐⭐)"
            elif predicted_rating >= 3.5:
                rating_desc = "Good (⭐⭐⭐)"
            else:
                rating_desc = "Average (⭐⭐)"
                
            print(f"   🎯 Predicted Rating: {predicted_rating}/5.0 - {rating_desc}")
            print(f"   📊 Confidence: {confidence}")
    
    # Model insights
    print("\\n" + "=" * 50)
    print("🧠 Model Insights:")
    print("-" * 50)
    print("🔍 Most Important Features:")
    print("   1. Author Identity (33.7% importance)")
    print("   2. Author Book Count (13.9% importance)")  
    print("   3. Reviews Count (11.8% importance)")
    print("   4. Ratings Count (11.0% importance)")
    
    print("\\n📈 Model Performance:")
    print("   • Test R² Score: 0.544 (Medium accuracy)")
    print("   • Mean Absolute Error: ±0.121 rating points")
    print("   • Training Data: 990+ books from Goodreads")
    
    print("\\n🎯 Prediction Guidelines:")
    print("   • Higher ratings count → More reliable prediction")
    print("   • Popular authors → Higher confidence")
    print("   • Typical rating range: 3.0 - 4.8")
    print("   • Average book rating: 4.1")
    
    # Interactive mode
    print("\\n" + "=" * 50)
    print("🎮 Try Your Own Prediction!")
    print("-" * 50)
    
    try:
        # Show available authors
        authors = predictor.get_top_authors()
        print("\\n📚 Available Authors (Top 20):")
        for i, author in enumerate(authors[:10], 1):
            print(f"   {i:2d}. {author}")
        print("   ... and 10 more authors")
        
        # Get user input
        print("\\nEnter details for your book prediction:")
        
        # Author selection
        author_input = input("\\n👤 Author name (from list above): ").strip()
        if not author_input:
            author_input = "Stephen King"  # Default
            
        # Ratings count
        try:
            ratings_input = input("📊 Expected ratings count (e.g., 50000): ").strip()
            ratings_count = int(ratings_input) if ratings_input else 50000
        except ValueError:
            ratings_count = 50000
            
        # Reviews count
        try:
            reviews_input = input("💬 Expected reviews count (optional): ").strip()
            reviews_count = int(reviews_input) if reviews_input else None
        except ValueError:
            reviews_count = None
        
        # Make prediction
        print("\\n🤔 Analyzing book characteristics...")
        result = predictor.predict_book_rating(author_input, ratings_count, reviews_count)
        
        if "error" in result:
            print(f"❌ Error: {result['error']}")
        else:
            predicted_rating = result["predicted_rating"]
            confidence = result["confidence"]
            
            print(f"\\n🎯 PREDICTION RESULT:")
            print(f"   📖 Book by: {author_input}")
            print(f"   ⭐ Predicted Rating: {predicted_rating}/5.0")
            print(f"   🎯 Confidence: {confidence}")
            
            if confidence == "High":
                print("   ✅ High confidence prediction!")
            elif confidence == "Medium":
                print("   ⚠️  Medium confidence - consider more popular author")
            else:
                print("   🔄 Low confidence - prediction may be less reliable")
                
    except KeyboardInterrupt:
        print("\\n\\n👋 Demo ended by user.")
    except Exception as e:
        print(f"\\n❌ Error during interactive prediction: {e}")
    
    print("\\n" + "=" * 50)
    print("🚀 Want to try the web interface?")
    print("   Run: streamlit run streamlit_app.py")
    print("   Then visit: http://localhost:8501")
    print("\\n📚 Thanks for trying Book Popularity Predictor!")
    print("=" * 50)

if __name__ == "__main__":
    run_demo()