This repository contains the code to test and compare the performance of geospatial computations with simple-data-analysis (Deno), GeoPandas (Python) and sf (R).

Each script has been run ten times on a MacBook Pro (Apple M1 Pro / 16 GB).

Here are the steps taken in each script:

• Loading a CSV file with the latitude and longitude of all trees in the city of Montreal (128 MB / 316,321 rows / 33 columns).
• Creating point geometries from the latitude and longitude columns.
• Loading the Montreal neighbourhoods as polygons (991 KB / 91 rows / 6 columns).
• Performing a spatial join to match each tree to its neighbourhood.
• Counting the number of trees in each neighbourhood.

The results are presented in this notebook.