Installation

Prerequisites

The project requires an installation of python 3.10 or newer. GprMax requires an installation of a C compiler like gcc. More info at http://docs.gprmax.com/en/latest/include_readme.html.

Quick install

By far the easiest way to install all the required dependencies for the project is to use the provided install.sh bash script found in the root directory of the repository.

First, clone the repository to your machine with:

git clone git@github.com:ThomasRigoni7/PINN4GPR.git

or using https with:

git clone https://github.com/ThomasRigoni7/PINN4GPR.git

then move inside the root directory:

cd PINN4GPR

create and activate a virtual environment with the venv packge.

*path/to/python3.10* -m venv .venv
source .venv/bin/activate

and execute the script:

./install.sh

This will:

• Clone a fork of gprMax from https://github.com/ThomasRigoni7/PINN4GPR into gprmax_repo

• Install all the required packages into the virtual environment, including building and installing gprMax.

The script will ask for a confirmation to install cuda support for gprMax:

Do you wish to install cuda support for gprMax? An existing cuda installation is required.
1) Yes
2) No
#?

This will install the pycuda package. This step might fail if no cuda installation is present on the system.

For future sessions, just activate the environment with the standard

source .venv/bin/activate

and deactivate with

deactivate

3D ballast simuation

The 3D ballast simulation module src/dataset_creation/ballast_simulation3D.py requires the installation of the chrono physics engine and the pychrono python bindings, which are not included in the standard installation.

More information on the creation of a conda environment for pychrono can be found here.