validate_examples.py

"""
Validation script for example outputs.
This script validates that the example scripts produce expected results.
"""

import pandas as pd
import sys
from pathlib import Path

# Add parent directory to path
sys.path.insert(0, str(Path(__file__).parent))

from batchguard import BatchGuard, min_max, total_range


def validate_synthetic_test_data():
    """Validate synthetic_test_data.py example."""
    print("Validating synthetic_test_data.py example...")
    
    transactions = pd.DataFrame({
        'account_id': ['A001', 'A001', 'A001', 'A002', 'A002', 'A002', 'A003', 'A003'],
        'transaction_type': ['withdrawal', 'withdrawal', 'withdrawal', 
                             'deposit', 'deposit', 'deposit', 'withdrawal', 'withdrawal'],
        'amount': [5000, 6000, 7000,  # Account A001: only large withdrawals
                   100, 150, 200,      # Account A002: normal deposits
                   8000, 9000],        # Account A003: only large withdrawals
    })
    
    guard = BatchGuard(transactions)
    guard.add_constraint(
        min_max('account_id', 'amount', min_val=50, max_val=2000)
    )
    
    fixed_transactions, report = guard.enforce()
    
    # Validate results
    assert all(50 <= val <= 2000 for val in fixed_transactions['amount'].values), \
        "All amounts should be between 50 and 2000"
    
    # Check that violations were resolved
    assert report.summary['satisfied'] == 1, "Constraint should be satisfied"
    assert report.summary['rows_modified'] == 5, "5 rows should be modified"
    assert report.summary['values_changed'] == 5, "5 values should be changed"
    
    print("✓ synthetic_test_data.py validation passed")
    return True


def validate_scenario_simulation():
    """Validate scenario_simulation.py example."""
    print("Validating scenario_simulation.py example...")
    
    scenarios = pd.DataFrame({
        'scenario_id': ['S1', 'S1', 'S1', 'S2', 'S2', 'S2', 'S3', 'S3'],
        'category': ['revenue', 'costs', 'profit', 'revenue', 'costs', 'profit',
                     'revenue', 'costs'],
        'value': [500000, -300000, 200000,  # Scenario S1: total = 400K (OK)
                  600000, -400000, 200000,   # Scenario S2: total = 400K (OK)
                  800000, -500000],          # Scenario S3: total = 300K (below min)
    })
    
    guard = BatchGuard(scenarios)
    guard.add_constraint(
        total_range('scenario_id', 'value', min_total=400000, max_total=1000000)
    )
    
    fixed_scenarios, report = guard.enforce()
    
    # Validate results
    totals = fixed_scenarios.groupby('scenario_id')['value'].sum()
    assert all(400000 <= total <= 1000000 for total in totals.values), \
        "All scenario totals should be between 400K and 1M"
    
    # Check that violations were resolved
    assert report.summary['satisfied'] == 1, "Constraint should be satisfied"
    assert report.summary['rows_modified'] == 2, "2 rows should be modified"
    assert report.summary['values_changed'] == 4, "4 values should be changed"
    
    print("✓ scenario_simulation.py validation passed")
    return True


def validate_demo_datasets():
    """Validate demo_datasets.py example."""
    print("Validating demo_datasets.py example...")
    
    products = pd.DataFrame({
        'category': ['Electronics', 'Electronics', 'Electronics', 'Electronics',
                     'Clothing', 'Clothing', 'Clothing',
                     'Books', 'Books'],
        'product_name': ['Laptop', 'Phone', 'Tablet', 'Headphones',
                         'Shirt', 'Pants', 'Shoes',
                         'Novel', 'Textbook'],
        'price': [1200, 800, 600, 150,    # Electronics: wide range (OK)
                  20, 30, 40,              # Clothing: all very cheap (unrealistic)
                  15, 200],                # Books: one very expensive (unrealistic)
        'stock': [10, 50, 30, 100,
                  200, 150, 80,
                  500, 5],
    })
    
    guard = BatchGuard(products)
    guard.add_constraint(
        min_max('category', 'price', min_val=25, max_val=1500)
    )
    guard.add_constraint(
        total_range('category', 'stock', min_total=50, max_total=1000)
    )
    
    fixed_products, report = guard.enforce()
    
    # Validate price constraints
    category_prices = fixed_products.groupby('category')['price']
    for category, prices in category_prices:
        assert all(25 <= price <= 1500 for price in prices.values), \
            f"All prices in {category} should be between 25 and 1500"
    
    # Validate stock constraints
    category_stocks = fixed_products.groupby('category')['stock'].sum()
    assert all(50 <= stock <= 1000 for stock in category_stocks.values), \
        "All category stock totals should be between 50 and 1000"
    
    # Check that constraints were satisfied
    assert report.summary['satisfied'] == 2, "Both constraints should be satisfied"
    assert report.summary['rows_modified'] == 2, "2 rows should be modified"
    assert report.summary['values_changed'] == 2, "2 values should be changed"
    
    print("✓ demo_datasets.py validation passed")
    return True


if __name__ == '__main__':
    print("=" * 60)
    print("Validating Example Scripts")
    print("=" * 60)
    
    try:
        validate_synthetic_test_data()
        validate_scenario_simulation()
        validate_demo_datasets()
        
        print("\n" + "=" * 60)
        print("All validations passed! ✓")
        print("=" * 60)
        sys.exit(0)
    except Exception as e:
        print(f"\n✗ Validation failed: {e}")
        import traceback
        traceback.print_exc()
        sys.exit(1)