Code and supporting data for an analysis of stream and lake chemistry data from "synoptic" (large-scale, snapshot) samples from across the US.
The code primarily runs 4 machine learning models on the data and compares feature importances and feature partial dependencies of those models.
Special thanks to Gabriel Cano for his help in starting the project.