import pandas as pd
import numpy as np
from datetime import datetime, timedelta

def generate_synthetic_data(num_farms=10, days=365):
    np.random.seed(42)  # For reproducibility

    # Generate farm IDs
    farm_ids = [f"Farm_{i:03d}" for i in range(1, num_farms + 1)]

    # Generate dates
    end_date = datetime.now().date()
    start_date = end_date - timedelta(days=days-1)
    date_range = pd.date_range(start=start_date, end=end_date, freq='D')

    data = []

    for farm_id in farm_ids:
        for date in date_range:
            # Generate feed composition data
            protein_content = np.random.uniform(14, 18)  # %
            fiber_content = np.random.uniform(17, 23)  # %
            energy_content = np.random.uniform(1.5, 1.8)  # Mcal/kg

            # Generate cattle health indicators
            body_condition_score = np.random.uniform(2.5, 4.5)
            somatic_cell_count = np.random.lognormal(mean=5, sigma=0.5)

            # Generate environmental conditions
            temperature = np.random.normal(15, 5)  # °C
            humidity = np.random.uniform(40, 80)  # %

            # Generate milk production
            base_production = np.random.uniform(20, 35)  # Liters per cow
            num_cows = np.random.randint(50, 500)

            # Add some seasonal variation
            seasonal_factor = 1 + 0.1 * np.sin(2 * np.pi * date.dayofyear / 365)

            # Calculate total milk production with some randomness
            milk_production = base_production * num_cows * seasonal_factor * np.random.uniform(0.9, 1.1)

            data.append({
                'farm_id': farm_id,
                'date': date,
                'protein_content': protein_content,
                'fiber_content': fiber_content,
                'energy_content': energy_content,
                'body_condition_score': body_condition_score,
                'somatic_cell_count': somatic_cell_count,
                'temperature': temperature,
                'humidity': humidity,
                'num_cows': num_cows,
                'milk_production': milk_production
            })

    df = pd.DataFrame(data)
    return df

# Generate the synthetic data
synthetic_data = generate_synthetic_data(num_farms=10, days=365)

# Display the first few rows and basic statistics
print(synthetic_data.head())
print(synthetic_data.describe())

# Save the data to a CSV file
synthetic_data.to_csv('dairy_farm_synthetic_data.csv', index=False)
print("Data saved to 'dairy_farm_synthetic_data.csv'")