fictional-pancake

Installation

We recommend the usage of mambaforge for setting up Python environment:

https://github.com/conda-forge/miniforge#mambaforge

## Install conda
mamba init

## Create new environment (named menv)
mamba env create --file=environment.yaml
mamba activate menv

Setup pre-commit hooks

pip install pre-commit
# Set up on the first usage
pre-commit install
pre-commit run --all-files

Run experiment

In order to run specific experiment override the experiment value accordingly:

# Resgated model working with directed graph data
PROJECT_ROOT=./ python src/train.py experiment=exp_resgated_digraph
# Resgated model working with multidirected graph data
PROJECT_ROOT=./ python src/train.py experiment=exp_resgated_multidigraph
# SIGN model working with directed graph data
PROJECT_ROOT=./ python src/train.py experiment=exp_sign_digraph

Transform data

Usually, we want to transform data before the experiments. dataset_name option must match the name of the dataset directory with graph data for transformation. Then you can run the following command:

# Transform multidigraph data using default transformations
PROJECT_ROOT=./ python src/transform.py graph=multidigraph dataset_name=random_species_10_01