Install dependencies

pip install numpy pandas scipy ripser scikit-learn faker networkx

Run the prototype

python main.py

Expected Output Example

===========================================================
Analyzing Dataset: E-commerce
===========================================================

1. Individual Table Analysis:
----------------------------------------

Analyzing table: customers
  Shape: (1000, 6)
  Dimension 0: 1 features
  Strong column relationships: customer_id-email, signup_date-country

Analyzing table: orders
  Shape: (5000, 5)
  Dimension 0: 1 features
  Dimension 1: 3 features
  Strong column relationships: order_id-customer_id

2. Table Relationship Discovery:
----------------------------------------

customers <-> orders:
  Confidence: 0.89
  Type: one-to-many
  Best join: customer_id = customer_id
  Join type: one-to-many

orders <-> order_items:
  Confidence: 0.95
  Type: one-to-many
  Best join: order_id = order_id
  Join type: one-to-many

3. Suggested Joins:
----------------------------------------

Join 1:
  Tables: customers + orders
  Confidence: 0.89
  SQL:
SELECT *
FROM customers t1
ONE_TO_MANY JOIN orders t2
    ON t1.customer_id = t2.customer_id