import pandas as pd
import numpy as np
print(pd.__version__)

rng = np.random.default_rng(seed=88)

regions = ['North', 'South', 'East', 'West']
customers = pd.DataFrame({
    'customer_id': range(1, 41),
    'region': rng.choice(regions, size=40),
    'signup_date': pd.date_range('2025-01-01', periods=40, freq='9D')
})
customers.head()

products = pd.DataFrame({
    'product_id': range(101, 121),
    'product_name': [f'Product {chr(65 + i % 26)}{i}' for i in range(20)],
    'category': rng.choice(['Electronics', 'Home', 'Apparel', 'Outdoors'], size=20),
    'price': rng.uniform(8, 250, size=20).round(2)
})
products.head()

n_orders = 500
order_rows = []
order_dates = pd.date_range('2026-01-01', periods=200, freq='D')

for order_id in range(20001, 20001 + n_orders):
    customer_id = int(rng.choice(customers['customer_id']))
    product_id = int(rng.choice(products['product_id']))
    quantity = int(rng.integers(1, 5))
    order_date = rng.choice(order_dates)
    order_rows.append([order_id, customer_id, product_id, quantity, order_date])

orders = pd.DataFrame(order_rows, columns=['order_id', 'customer_id', 'product_id', 'quantity', 'order_date'])
print(orders.shape)

customers.to_csv('customers.csv', index=False)
products.to_csv('products.csv', index=False)
orders.to_csv('orders.csv', index=False)
customers = pd.read_csv('customers.csv', parse_dates=['signup_date'])
products = pd.read_csv('products.csv')
orders = pd.read_csv('orders.csv', parse_dates=['order_date'])
orders.head()

print('customers:', customers.shape)
print('products:', products.shape)
print('orders:', orders.shape)

orders.info()

full_orders = orders.merge(customers, on='customer_id', how='left').merge(products, on='product_id', how='left')
full_orders['revenue'] = full_orders['quantity'] * full_orders['price']
full_orders[['order_id', 'region', 'product_name', 'category', 'quantity', 'price', 'revenue']].head()

print(full_orders.shape)
full_orders.isna().sum().sum()

top_products = full_orders.groupby('product_name')['revenue'].sum().sort_values(ascending=False).head(5)
top_products.round(2)

customer_totals = full_orders.groupby(['customer_id', 'region'])['revenue'].agg(['sum', 'count']).reset_index()
customer_totals.columns = ['customer_id', 'region', 'total_spent', 'order_count']
top_customers = customer_totals.sort_values('total_spent', ascending=False).head(5)
top_customers.round(2)

def segment(row):
    if row['order_count'] >= 15 and row['total_spent'] >= 800:
        return 'VIP'
    elif row['order_count'] >= 5:
        return 'Regular'
    else:
        return 'Occasional'

customer_totals['segment'] = customer_totals.apply(segment, axis=1)
customer_totals['segment'].value_counts()

segment_summary = customer_totals.groupby('segment')['total_spent'].agg(['mean', 'sum']).round(2)
segment_summary

region_summary = full_orders.groupby('region').agg(
    total_revenue=('revenue', 'sum'),
    orders=('order_id', 'count'),
    avg_order_value=('revenue', 'mean')
).round(2).sort_values('total_revenue', ascending=False)
region_summary

import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(8, 5))
top_products.plot(kind='barh', ax=ax, color='steelblue')
ax.set_title('Top 5 Products by Revenue')
ax.set_xlabel('Revenue ($)')
ax.invert_yaxis()
plt.tight_layout()
plt.savefig('top_products.png', dpi=150)
plt.close(fig)
print('Saved top_products.png')

fig, ax = plt.subplots(figsize=(7, 5))
customer_totals['segment'].value_counts().reindex(['VIP', 'Regular', 'Occasional']).plot(kind='bar', ax=ax, color=['gold', 'steelblue', 'lightgray'])
ax.set_title('Customers by Segment')
ax.set_ylabel('Number of Customers')
plt.xticks(rotation=0)
plt.tight_layout()
plt.savefig('customer_segments.png', dpi=150)
plt.close(fig)
print('Saved customer_segments.png')