Travelling Salesman Problem in Snowflake

1. Overview

This example demonstrates how to use Snowflake's User-Defind-Table-Functions (UDTFs) to solve the Travelling Salesman Problem. The UDTF makes use of a library called NetworkX which includes many graph algorithms that can be used in Snowflake's UDTFs.

2. Prerequisites

• Snowflake Account
• Snowpark for Python
• The examples also use the following Python libraries:

  networkx
  pandas
  matplotlib
  urllib
  requests
  geojson
  gmaps
  

• Jupyter or JupyterLab

3. What you'll learn

• How to use Snowpark for Python to create User-Defined-Table-Functions (UDTFs)
• How to do graph analysis with Snowflake UDTFs

4. Usage/Steps

1. Open terminal and clone this repo.

   git clone https://github.com/Snowflake-Labs/snowpark-python-examples

2. Update creds.json to reflect your snowflake environment.

3. Change to the Graph Analysis directory and launch JupyterLab (or Juypter)

   jupyter lab

4. Paste the URL in a browser window and once JupyterLab comes up.

5. Run the cells of the Notebook.