Recently, I thought back to a few years ago, when I tried to process a large geospatial dataset with Python. You can only guess how it ended. My laptop refused to cooperate and froze spectacularly without failing.

Frustration ensued.

Fast forward today, I was experimenting with RAPIDS AI Suite and came across the same dataset. I immediately knew what to do. So I jumped into coding.

The RAPIDS suite of open source software libraries and APIs gives you the ability to execute end-to-end data science and analytics pipelines entirely on GPUs.

In this tutorial, I will go through a complete Geospatial data analysis example with cuDF and cuSpatial libraries. With spatial data, most often, we associate relationships of objects in space to one another, and therefore, the spatial join is crucial in many GIS applications.

In this example, I’ll show you how to perform a spatial join with big data efficiently using GPU to speed up the process. The whole process of reading and executing a spatial join on 8 million rows data takes less than a minute with GPU processing.

#data-science #python #geospatial #rapids-ai #gis