analysis.py

import pandas as pd
import matplotlib.pyplot as plt

# Load the dataset
# Download from: https://www.kaggle.com/datasets/carrie1/ecommerce-data
# Or use: https://archive.ics.uci.edu/ml/datasets/Online+Retail
df = pd.read_csv('online_retail.csv', encoding='ISO-8859-1')

print()
print(f"Dataset loaded: {len(df)} rows")
print()

# Data Cleaning Start
print("Data Cleaning...")
print()

# Remove rows with missing CustomerID or Description
initial_rows = len(df)
df = df.dropna(subset=['CustomerID', 'Description'])
print(f"Removed {initial_rows - len(df)} rows with missing data")

# Remove cancelled orders - orders starting with C
df = df[~df['InvoiceNo'].astype(str).str.startswith('C')]
print(f"Remaining rows after removing cancellations: {len(df)}")

# Remove rows with negative or zero quantities
df = df[(df['Quantity'] > 0) & (df['UnitPrice'] > 0)]
print(f"Remaining rows after removing invalid quantities/prices: {len(df)}")

# Calculate total revenue
df['TotalRevenue'] = df['Quantity'] * df['UnitPrice']

# Convert InvoiceDate to datetime for python understanding
df['InvoiceDate'] = pd.to_datetime(df['InvoiceDate'])

# Extract month and year
df['Month'] = df['InvoiceDate'].dt.month
df['Year'] = df['InvoiceDate'].dt.year
df['YearMonth'] = df['InvoiceDate'].dt.to_period('M')

print("\nData cleaning complete...")
print(f"Final dataset: {len(df)} rows")

# QUESTION 1: What are the top 5 best-selling product categories by revenue?

print()
print("Question 1: Top 5 Best-Selling Products by Total Revenue")
print()

# Group by product description and calculate total revenue
product_revenue = df.groupby('Description').agg({
    'TotalRevenue': 'sum',
    'Quantity': 'sum',
    'InvoiceNo': 'count'
}).reset_index()

product_revenue.columns = ['Product', 'TotalRevenue', 'TotalQuantity', 'OrderCount']

# Sort by revenue and get top 5
top_5_products = product_revenue.sort_values('TotalRevenue', ascending=False).head(5)

print("\nTop 5 Products by Revenue:")
print(top_5_products.to_string(index=False))

print("\nAnalysis:")
print(f"1. Best-selling product: {top_5_products.iloc[0]['Product']}")
print(f"   - Revenue: £{top_5_products.iloc[0]['TotalRevenue']:,.2f}")
print(f"   - Quantity sold: {top_5_products.iloc[0]['TotalQuantity']:,.0f} units")
print(f"\n2. The top 5 products generated £{top_5_products['TotalRevenue'].sum():,.2f} in total revenue")

# QUESTION 2: What are the monthly sales trends?

print()
print("Question 2: Monthly Sales Trends")
print()

# Group by month and calculate total revenue
monthly_sales = df.groupby('YearMonth').agg({
    'TotalRevenue': 'sum',
    'InvoiceNo': 'nunique'
}).reset_index()

monthly_sales.columns = ['Month', 'TotalRevenue', 'OrderCount']

# Sort by revenue to find best/worst months
monthly_sales_sorted = monthly_sales.sort_values('TotalRevenue', ascending=False)

print("\nMonthly Sales Summary:")
print(monthly_sales.to_string(index=False))

print("\nAnalysis:")
print(f"1. Best month: {monthly_sales_sorted.iloc[0]['Month']}")
print(f"   - Revenue: £{monthly_sales_sorted.iloc[0]['TotalRevenue']:,.2f}")
print(f"   - Orders: {monthly_sales_sorted.iloc[0]['OrderCount']:,.0f}")

print(f"\n2. Worst month: {monthly_sales_sorted.iloc[-1]['Month']}")
print(f"   - Revenue: £{monthly_sales_sorted.iloc[-1]['TotalRevenue']:,.2f}")
print(f"   - Orders: {monthly_sales_sorted.iloc[-1]['OrderCount']:,.0f}")

avg_monthly_revenue = monthly_sales['TotalRevenue'].mean()
print(f"\n3. Average monthly revenue: £{avg_monthly_revenue:,.2f}")

# Visual (Stretch Challenge)

print()
print("Creating data visualization...")
print()

# Create a figure with two subplots
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 6))

# Chart 1: Top 5 Products by Revenue
ax1.barh(top_5_products['Product'], top_5_products['TotalRevenue'], color='steelblue')
ax1.set_xlabel('Total Revenue (£)', fontsize=12)
ax1.set_ylabel('Product', fontsize=12)
ax1.set_title('Top 5 Best-Selling Products by Revenue', fontsize=14, fontweight='bold')
ax1.invert_yaxis()

# Add revenue labels on bars
for i, v in enumerate(top_5_products['TotalRevenue']):
    ax1.text(v, i, f' £{v:,.0f}', va='center', fontsize=10)

# Chart 2: Monthly Sales Trend
monthly_sales['MonthStr'] = monthly_sales['Month'].astype(str)
ax2.plot(monthly_sales['MonthStr'], monthly_sales['TotalRevenue'], 
         marker='o', linewidth=2, markersize=6, color='darkorange')
ax2.set_xlabel('Month', fontsize=12)
ax2.set_ylabel('Total Revenue (£)', fontsize=12)
ax2.set_title('Monthly Sales Trend', fontsize=14, fontweight='bold')
ax2.grid(True, alpha=0.3)
ax2.tick_params(axis='x', rotation=45)

# Format y-axis to show currency
ax2.yaxis.set_major_formatter(plt.FuncFormatter(lambda x, p: f'£{x/1000:.0f}K'))

# Save figure
plt.tight_layout()
plt.savefig('sales_analysis_results.png', dpi=300, bbox_inches='tight')
print("Visualization saved as 'sales_analysis_results.png'")

plt.show()

print()
print("Visualization Completed")