spagnoloG / WAMF-FPI-open

Opensource implementation of WAMF-FPI model, using pytorch lighning and w&b integration

WAMF-FPI-open

Open Source Implementation of the WAMF-FPI algorithm in Python. Including additional features and continious improvements.

Start (OldSchool way) ~ Not recommended

sudo add-apt-repository ppa:ubuntugis/ppa -y && sudo apt update -y
sudo apt install gdal-bin libgdal-dev -y  # Hopefully this will work
sudo apt install python3.11 python3.11-venv git  -y
export CPLUS_INCLUDE_PATH=/usr/include/gdal
export C_INCLUDE_PATH=/usr/include/gdal
pip install GDAL --global-option=build_ext --global-option="-I/usr/include/gdal" --global-option="-lgdal"
git clone https://github.com/spagnoloG/WAMF-FPI-open.git
cd WAMF-FPI-open
python3 -m venv .venv
. .venv/bin/activate
pip install -r requirements.txt

Start (nix way) ~ Ubuntu

• Install nix package manager

sudo apt install curl -y
sh <(curl -L https://nixos.org/nix/install) --daemon

• Enable nix-flakes

echo "experimental-features = nix-command flakes" | sudo tee -a /etc/nix/nix.conf > /dev/null

• Clone the repo and enter development environment

nix develop github:spagnoloG/WAMF-FPI-open

Start ~ NixOS

nix develop github:spagnoloG/WAMF-FPI-open

After setting up

To automatically enter the environment when you cd into project, run the following command:

sudo apt install -y direnv # or nix-shell -p direnv
direnv allow

Train

python code/main.py fit --conf code/conf/config.yaml # Modify conf based on your requirements

Predict from checkpoint

python code/main.py predict --conf code/conf/config.yaml --ckpt_path=<path>