#!/usr/bin/env python3
"""
Example script demonstrating how to load the IFVI Value Factors dataset
in different formats (JSON, CSV, Parquet).
"""

import os
import json
import pandas as pd
import matplotlib.pyplot as plt
from pathlib import Path

# Get the repository root directory
REPO_ROOT = Path(__file__).parent.parent.parent.absolute()

def load_json_data(country_name):
    """
    Load value factors for a specific country from JSON file.
    
    Args:
        country_name (str): Name of the country
        
    Returns:
        dict: Value factors data for the country
    """
    # First try to find the country in the by-territory structure
    continents = ["Africa", "Asia", "Europe", "North America", "Oceania", "South America"]
    
    for continent in continents:
        country_path = REPO_ROOT / "data" / "by-territory" / "by-continent" / continent / f"{country_name}.json"
        if country_path.exists():
            # This is a Git LFS file, so we need to handle it differently
            # In a real implementation, you would use the git-lfs Python package
            # For this example, we'll assume the file is already pulled
            
            # For demonstration purposes, we'll load the composite data instead
            composite_path = REPO_ROOT / "data" / "composite-data" / "all-formats" / "composite_value_factors.json"
            if composite_path.exists():
                with open(composite_path, 'r') as f:
                    data = json.load(f)
                # Filter for the specific country
                if country_name in data:
                    return data[country_name]
            
            print(f"Country file found but using composite data instead: {country_path}")
            return {}
    
    print(f"Country not found: {country_name}")
    return {}

def load_csv_data():
    """
    Load the entire dataset from CSV.
    
    Returns:
        pandas.DataFrame: Value factors data
    """
    csv_path = REPO_ROOT / "data" / "composite-data" / "all-formats" / "composite_value_factors.csv"
    if csv_path.exists():
        return pd.read_csv(csv_path)
    else:
        print(f"CSV file not found: {csv_path}")
        return pd.DataFrame()

def load_parquet_data():
    """
    Load the entire dataset from Parquet.
    
    Returns:
        pandas.DataFrame: Value factors data
    """
    parquet_path = REPO_ROOT / "data" / "composite-data" / "all-formats" / "composite_value_factors.parquet"
    if parquet_path.exists():
        return pd.read_parquet(parquet_path)
    else:
        print(f"Parquet file not found: {parquet_path}")
        return pd.DataFrame()

def compare_countries(countries, category, impact):
    """
    Compare value factors across countries for a specific category and impact.
    
    Args:
        countries (list): List of country names
        category (str): Environmental impact category
        impact (str): Impact type
        
    Returns:
        pandas.DataFrame: Comparison data
    """
    data = load_csv_data()
    if data.empty:
        return pd.DataFrame()
    
    # Filter for the specific category and impact
    filtered_data = data[(data['Category'] == category) & (data['Impact'] == impact)]
    
    # Filter for the specified countries
    country_data = filtered_data[filtered_data['Territory'].isin(countries)]
    
    return country_data

def plot_country_comparison(countries, category, impact):
    """
    Plot a comparison of value factors across countries.
    
    Args:
        countries (list): List of country names
        category (str): Environmental impact category
        impact (str): Impact type
    """
    comparison_data = compare_countries(countries, category, impact)
    if comparison_data.empty:
        print("No data available for the specified parameters")
        return
    
    plt.figure(figsize=(10, 6))
    plt.bar(comparison_data['Territory'], comparison_data['ValueFactor'])
    plt.title(f"{category} - {impact} Value Factors by Country")
    plt.xlabel("Country")
    plt.ylabel("Value Factor (USD)")
    plt.xticks(rotation=45)
    plt.tight_layout()
    
    # Save the plot
    output_dir = REPO_ROOT / "examples" / "outputs"
    output_dir.mkdir(parents=True, exist_ok=True)
    plt.savefig(output_dir / f"{category}_{impact}_comparison.png")
    plt.close()

def main():
    """Main function demonstrating the use of the dataset."""
    # Example 1: Load data for a specific country
    country_data = load_json_data("United States")
    print(f"Number of value factors for United States: {len(country_data) if country_data else 0}")
    
    # Example 2: Load CSV data
    csv_data = load_csv_data()
    print(f"CSV data shape: {csv_data.shape}")
    
    # Example 3: Load Parquet data
    parquet_data = load_parquet_data()
    print(f"Parquet data shape: {parquet_data.shape}")
    
    # Example 4: Compare countries
    countries_to_compare = ["United States", "China", "Germany", "Brazil", "India"]
    comparison = compare_countries(countries_to_compare, "PM2.5", "Primary Health")
    print("\nComparison of PM2.5 Primary Health value factors:")
    print(comparison[['Territory', 'Location', 'ValueFactor']].to_string(index=False))
    
    # Example 5: Plot comparison
    plot_country_comparison(countries_to_compare, "PM2.5", "Primary Health")
    print("\nPlot saved to examples/outputs/PM2.5_Primary Health_comparison.png")

if __name__ == "__main__":
    main()