3.1. Capstone Project 1

Capstone Project 1 - Mind Map

Google colab notebook with Python code

https://colab.research.google.com/drive/16Gi_HjxdSBGkMvZsyEWUDD8d-1qn8VOH?usp=sharing

Explaination of the code by employing ChatGPT (just the first part)

📘 Project Overview: Product Value Analysis with Pareto Charts

This notebook performs an exploratory analysis on an e-commerce dataset, focusing on identifying the most valuable product shipping and customer locations using the Pareto principle (80/20 rule).

✅ Key Steps Performed

1. Data Loading
   
   

   • Reads a CSV file hosted on Google Drive into a pandas DataFrame.
     
     

2. Data Inspection
   
   

   • Displays the structure and column names of the dataset using df.columns and df.info().
     
     

3. Derived Feature: Weight per Unit
   
   

   • A new feature weight_per_unit is computed by dividing the total weight (WeightKg) by the number of items (ItemCount) per order.
     
     

4. Pareto Analysis Function
   
   

   • A custom function performs Pareto analysis by:
     
     

     • Grouping the data by a selected category (e.g., shipping country or customer country).
       
       

     • Summing the selected value field (e.g., Price).
       
       

     • Sorting and computing the cumulative percentage.
       
       

     • Plotting a Pareto chart: bar plot for values and line plot for cumulative percentage.
       
       

     • Drawing a horizontal line at 80% to visually identify the most impactful categories.
       
       

5. Applications of the Analysis
   
   

   • The Pareto function is applied twice:
     
     

     • Once to evaluate revenue distribution by shipping country.
       
       

     • Once to evaluate revenue distribution by customer country.
       
       

🔍 Core Concepts Employed

• Pandas Data Manipulation: grouping, sorting, column creation.
  
  

• Pareto Principle (80/20 Rule): a small number of categories often account for the majority of the effect (e.g., sales).
  
  

• Matplotlib Visualization: custom plots combining bars and lines for business insights.
  
  

• Feature Engineering: deriving new data features like weight per unit to aid in analysis.