This repo contains the experiments for our paper "Flexible and Efficient Probabilistic PDE Solvers through Gaussian Markov Random Fields".
-
Clone this repo.
-
In the root directory of this project, run
julia --project=.followed by] dev .
Please note that GaussianMarkovRandomFields.jl (the main dependency for these experiments) is not registered yet.
Until it is, you can install it directly from GitHub through
] add https://github.com/timweiland/GaussianMarkovRandomFields.jl.
./hpc contains utilities to get this code running in a singularity container for HPC.
Once you've figured this out, you can grab the data from the sources specified in the subdirectories of ./data.
Afterwards, you can run the scripts in ./scripts.