Spatial Joins at Scale: Unlocking Advanced Geospatial Analytics with Wherobots

In this session, you'll learn how to seamlessly integrate Python and Wherobots to perform advanced spatial joins and analyses on geospatial data. Webinar Highlights: - Learn how to join datasets using spatial predicates like ST_Intersects to combine facilities with administrative boundaries—ensuring only relevant spatial relationships are processed. - See how to leverage the powerful ST_AKNN function to efficiently find the k-nearest neighbors for each facility. We’ll explore a Pythonic approach to identify the closest centroids to your facilities, ensuring a highly optimized query without the overhead of expensive cross joins. - Understand how to apply spatial filters to extract a targeted subset of your data. We'll walk through filtering facilities that fall within a specified geographic boundary. - Explore strategies like repartitioning by geohash to improve performance and reduce data shuffling during spatial joins, making your spatial queries scalable for large datasets. - Finally, learn how to visualize your spatial join results interactively using SedonaKepler and SedonaPyDeck, turning complex geospatial data into actionable insights with rich, interactive maps. Timestamps 00:00 Introduction 03:01 Introduction to Wherobots, Apache Sedona and product architecture 08:16 Performance benchmarks: Wherobots vs. Sedona, BigQuery and Snowflake 09:55 Create account and get started with Wherobots | Start a notebook, setting runtime, adding Python libraries, cloud storage, and spatial catalog 18:40 Start demo | Connect to runtime and notebook, importing functions from PySpark 20:48 Load data | Polygons from Overture Maps and points from Foursquare Open Places Dataset 23:49 Standard Spatial Join (Pythonic Approach) 27:32 Spatial aggregate | Efficiently count points in each polygon 29:13 K Nearest Neighbor Join 37:22 Advanced optimization techniques | Cluster data using Geohash 47:28 Visualize Spatial Join results with SedonaKepler and SedonaPyDeck 48:11 Summary and Q&A Get started with Wherobots: https://bit.ly/3Q9PcPe