import pandas as pd
import numpy as np

# Set a random seed for reproducibility
np.random.seed(42)

# Generate data for 100 houses
num_rows = 100

# Size: 800-3000 square feet
size = np.random.randint(800, 3001, size=num_rows)

# Price: Linear relationship with size (price = 200 * size + 50000 + noise)
noise = np.random.normal(0, 20000, size=num_rows)
price = 200 * size + 50000 + noise

# Create DataFrame
df = pd.DataFrame({
    'size': size,
    'price': price
})

# Save to CSV
df.to_csv("house_prices.csv", index=False)
print("Generated house_prices.csv with 100 rows!")