Machine Learning and BigQuery GIS

This lab demonstrates how to perform a clustering analysis in BigQuery GIS using Python and Jupyter notebooks. The lab leverages the built-in DBScan clustering function in BigQuery GIS to cluster street trees in San Francisco from the Google public datasets. It also shows how to analyze the cluster results using the pandas library within Python.

Learning Objectives

Upon completion of this lab, you will be able to:

• Interact with BigQuery GIS datasets within Jupyter notebooks
• Perform a clustering analysis on GEOGRAPHY data
• Use pandas to analyze cluster analysis results

Intended Audience

This lab is intended for:

• GIS engineers
• Data engineers dealing with location-based data
• Developers looking to leverage geospatial information
• Data scientists leveraging geospatial data

Prerequisites

You should possess:

• Basic understanding of relational databases and ANSI SQL
• Basic understanding of Python
• Familiarity with BigQuery GIS's GEOGRAPHY datatype

Updates

August 20th, 2025 - Resvoled issue connecting to the lab's notebook

June 24th, 2025 - Updated GCP VM instance image

May 27th, 2025 - Migrated lab environment, updated lab permissions, and cleared Jupyter Notebook outputs

August 7th, 2024 - Added password protection to notebook