data_loader.py

import pandas as pd
import numpy as np
from functools import lru_cache

@lru_cache(maxsize=1)
def load_movie_data(file_path="movie_metadata.parquet"):
    """
    Loads and caches the movie metadata parquet file.
    Performs data cleaning, calculates the trending score, and creates standard fields.
    """
    try:
        df = pd.read_parquet(file_path)
    except Exception as e:
        print(f"Error loading {file_path}: {e}")
        # Return a fallback empty DataFrame with correct columns if load fails
        df = pd.DataFrame(columns=["movie_id", "title", "year", "avg_rating", "rating_count"])
    
    # Data cleaning: fill missing values
    df["year"] = df["year"].fillna(0).astype(int)
    df["avg_rating"] = df["avg_rating"].fillna(0.0).astype(float)
    df["rating_count"] = df["rating_count"].fillna(0).astype(int)
    df["title"] = df["title"].fillna("Unknown Title").astype(str)
    
    # Calculate trending score: (avg_rating * log(rating_count))
    # Add a small epsilon to rating_count to prevent log(0)
    df["trending_score"] = df["avg_rating"] * np.log(df["rating_count"] + 1)
    
    return df

def get_trending_movies(df, limit=20):
    """
    Returns the top N movies sorted by trending score descending.
    """
    return df.sort_values(by="trending_score", ascending=False).head(limit)

def get_kpi_statistics(df):
    """
    Calculates key metrics for the analytics dashboard:
    - Total Movies
    - Average Rating
    - Average Votes (Rating Count)
    - Most Popular Movie (Title & Count)
    - Newest Movie (Title & Year)
    """
    if df.empty:
        return {
            "total_movies": 0,
            "avg_rating": 0.0,
            "avg_votes": 0,
            "most_popular": "N/A",
            "newest_movie": "N/A"
        }
        
    total_movies = len(df)
    avg_rating = float(df["avg_rating"].mean())
    avg_votes = int(df["rating_count"].mean())
    
    # Most popular movie (highest rating_count)
    pop_idx = df["rating_count"].idxmax()
    most_popular = f"{df.loc[pop_idx, 'title']} ({df.loc[pop_idx, 'rating_count']:,} votes)"
    
    # Newest movie (highest year, filtering out 0 years)
    valid_years_df = df[df["year"] > 0]
    if not valid_years_df.empty:
        new_idx = valid_years_df["year"].idxmax()
        newest_movie = f"{valid_years_df.loc[new_idx, 'title']} ({valid_years_df.loc[new_idx, 'year']})"
    else:
        newest_movie = "N/A"
        
    return {
        "total_movies": total_movies,
        "avg_rating": avg_rating,
        "avg_votes": avg_votes,
        "most_popular": most_popular,
        "newest_movie": newest_movie
    }