hf_usage_example.py

#!/usr/bin/env python3
"""
Example usage of the agricultural data loader with Hugging Face integration.
Shows different ways to load and use the data.
"""

import os
import warnings
warnings.filterwarnings('ignore')

from data_loader import AgriculturalDataLoader
from analysis_tools import AgriculturalAnalyzer

def example_local_usage():
    """Example: Load from local files."""
    print("📁 EXAMPLE 1: Loading from local files")
    print("-" * 40)
    
    # Create loader for local files
    loader = AgriculturalDataLoader.create_local_loader(
        data_path="/Users/tracyandre/Downloads/OneDrive_1_9-17-2025"
    )
    
    # Load and analyze data
    df = loader.load_all_files()
    print(f"✅ Loaded {len(df):,} records from local files")
    
    # Basic analysis
    analyzer = AgriculturalAnalyzer(loader)
    trends = analyzer.analyze_weed_pressure_trends()
    print(f"📊 Average IFT: {trends['summary']['mean_ift']:.2f}")
    
    return df

def example_hf_usage():
    """Example: Load from Hugging Face (if available)."""
    print("\n🤗 EXAMPLE 2: Loading from Hugging Face")
    print("-" * 40)
    
    # Check if HF token is available
    if not os.environ.get("HF_TOKEN"):
        print("⚠️  No HF_TOKEN found - skipping HF example")
        print("💡 Set HF_TOKEN environment variable to use this feature")
        return None
    
    try:
        # Create loader for Hugging Face
        loader = AgriculturalDataLoader.create_hf_loader(
            dataset_id="HackathonCRA/2024"
        )
        
        # Load and analyze data
        df = loader.load_all_files()
        print(f"✅ Loaded {len(df):,} records from Hugging Face")
        
        # Basic analysis
        analyzer = AgriculturalAnalyzer(loader)
        trends = analyzer.analyze_weed_pressure_trends()
        print(f"📊 Average IFT: {trends['summary']['mean_ift']:.2f}")
        
        return df
        
    except Exception as e:
        print(f"❌ Failed to load from Hugging Face: {e}")
        return None

def example_automatic_fallback():
    """Example: Automatic fallback from HF to local."""
    print("\n🔄 EXAMPLE 3: Automatic fallback")
    print("-" * 40)
    
    # Create loader with HF preferred but local fallback
    loader = AgriculturalDataLoader(
        data_path="/Users/tracyandre/Downloads/OneDrive_1_9-17-2025",
        dataset_id="HackathonCRA/2024",
        use_hf=True  # Try HF first
    )
    
    # This will try HF first, then fallback to local if needed
    df = loader.load_all_files()
    print(f"✅ Loaded {len(df):,} records (with automatic source selection)")
    
    return df

def example_dynamic_switching():
    """Example: Dynamic switching between sources."""
    print("\n🔀 EXAMPLE 4: Dynamic source switching")
    print("-" * 40)
    
    # Create loader
    loader = AgriculturalDataLoader(
        data_path="/Users/tracyandre/Downloads/OneDrive_1_9-17-2025",
        dataset_id="HackathonCRA/2024"
    )
    
    # Load from local first
    loader.set_data_source(use_hf=False)
    df_local = loader.load_all_files()
    print(f"📁 Local source: {len(df_local):,} records")
    
    # Switch to HF (if available)
    if os.environ.get("HF_TOKEN"):
        try:
            loader.set_data_source(use_hf=True)
            df_hf = loader.load_all_files()
            print(f"🤗 HF source: {len(df_hf):,} records")
            
            # Compare
            if len(df_local) == len(df_hf):
                print("✅ Data consistency verified")
            else:
                print(f"⚠️  Data mismatch: {abs(len(df_local) - len(df_hf))} record difference")
                
        except Exception as e:
            print(f"🤗 HF switching failed: {e}")
    else:
        print("⚠️  No HF_TOKEN - skipping HF switch test")
    
    return df_local

def example_production_deployment():
    """Example: Production deployment configuration."""
    print("\n🚀 EXAMPLE 5: Production deployment setup")
    print("-" * 40)
    
    # Production configuration
    # This is how you'd set it up for Hugging Face Spaces deployment
    
    print("💡 For Hugging Face Spaces deployment:")
    print("1. Set HF_TOKEN as a Space secret")
    print("2. Configure the loader as follows:")
    print()
    
    config_code = '''
# In your app.py or gradio_app.py
import os
from data_loader import AgriculturalDataLoader

# Production configuration
hf_token = os.environ.get("HF_TOKEN")
dataset_id = "HackathonCRA/2024"

if hf_token:
    # Use HF dataset in production
    data_loader = AgriculturalDataLoader.create_hf_loader(
        dataset_id=dataset_id,
        hf_token=hf_token
    )
    print("🤗 Using Hugging Face dataset")
else:
    # Fallback for local development
    data_loader = AgriculturalDataLoader.create_local_loader(
        data_path="./data"  # Local data directory
    )
    print("📁 Using local files")
'''
    
    print(config_code)
    
    # Example of actual production setup
    try:
        hf_token = os.environ.get("HF_TOKEN")
        if hf_token:
            loader = AgriculturalDataLoader.create_hf_loader("HackathonCRA/2024", hf_token)
            print("✅ Production setup: HF dataset configured")
        else:
            loader = AgriculturalDataLoader.create_local_loader("/Users/tracyandre/Downloads/OneDrive_1_9-17-2025")
            print("✅ Development setup: Local files configured")
            
        df = loader.load_all_files()
        print(f"📊 Ready for production: {len(df):,} records available")
        
    except Exception as e:
        print(f"❌ Production setup failed: {e}")

def main():
    """Run all examples."""
    print("🚜 AGRICULTURAL DATA LOADER - USAGE EXAMPLES")
    print("=" * 60)
    
    # Run examples
    example_local_usage()
    example_hf_usage()
    example_automatic_fallback()
    example_dynamic_switching()
    example_production_deployment()
    
    print("\n" + "=" * 60)
    print("🎯 SUMMARY")
    print("=" * 60)
    print("""
The AgriculturalDataLoader now supports:

✅ Local file loading (CSV/Excel)
✅ Hugging Face dataset loading
✅ Automatic fallback (HF → Local)
✅ Dynamic source switching
✅ Production deployment ready

Key benefits:
🔄 Seamless data source switching
🚀 Cloud deployment ready
📊 Same analysis tools work with both sources
🔧 Easy configuration management
    """)
    
    print("🛠️  Next steps:")
    print("1. Upload your dataset to Hugging Face Hub")
    print("2. Set HF_TOKEN environment variable")
    print("3. Deploy to Hugging Face Spaces")
    print("4. Enjoy cloud-based agricultural analysis!")

if __name__ == "__main__":
    main()