Scaffolds Reading Group

Reading List

link to list

dPASP

Geh, R.L., Gonçalves, J., Silveira, I.C., Mauá, D.D. and Cozman, F.G., 2023. dPASP: a comprehensive differentiable probabilistic answer set programming environment for neurosymbolic learning and reasoning. arXiv preprint arXiv:2308.02944.

Installing dPASP

The GitHub repo is at https://github.com/kamel-usp/dpasp .

The tutorial link is broken, so I raised an issue here kamel-usp/dpasp#11 However the page is available via the Internet Archive at: https://web.archive.org/web/20240620194257/http:/kamel.ime.usp.br/pages/learn_dpasp

It seems that the software only works on Linux (not Mac) on x86 processors. Here are the steps I @RobBlackwell ) took to run it on an Azure Linux VM:

sudo apt install python3-venv build-essential

sudo add-apt-repository ppa:potassco/stable
sudo apt update
sudo apt install clingo libclingo-dev libncurses-dev

pip install --upgrade setuptools pip

pip install pasp-plp

dPASP via Docker on Apple Silicon

In the docker subdirectory, there is a Dockerfile that provides a working dPASP environment via Docker, even on Apple Silicon machines. I assume you have a working Docker (See Get Docker), and Make (probably via homebrew).

To build the Docker image (this may take some minutes!):

make build

To run the Docker image:

make run

The latter drops you into a shell prompt. You can test pasp by typing:

pasp --help

cd /app
pasp smokers.plp

I (@RobBlackwell) have tested this on Apple Silicon. It might work on Windows too?