This repository contains the source code used to retrieve and preprocess global satellite wildfire data (predominantly in Python), as well as to perform statistical analysis on it (in R), for my Mathematical Project in partial fulfillment of my degree.
The utils/ folder comprises quick and dirty Python scripts that I wrote to retrieve huge amounts of data (~200 GB) online.
The preprocess.R script processes various .nc files retrieved and turns it into a .csv file that is read by model_fitting.R to fit our Poisson hurdle model with latent spatial effects.
model_checking.R produces several model diagnostics and plots to examine the validity our model, and simulation.R performs a simulation study (not included in the report) using the parameters found in the model.
