Quantum algorithms for solving ASP programs.
I strongly recommend using a Python virtual environment to install all the required dependencies, in order to avoid (possibly) messing with other Python projects you may be working with.
A new venv can be created by running the following command:
python -m venv ./.py-venv --prompt qasp-solverTo enter (and then quit) the environment, the following commands can be used from the directory where the venv folder was created:
source ./.py-venv/bin/activate # Enter the venv
deactivate # Quit the venvThe rest of this section assumes that you run all the provided commands while inside the virtual environment.
First of all, install the Python dependencies specified in the requirements.txt file:
pip install -r ./requirements.txtAt the time of writing, the version of the tweedledum library packaged by PyPI is broken, thus it cannot be installed with pip. Instead, use the following commands to download, build, and install it from the publicly available sources:
# Install build dependencies (Ubuntu)
sudo apt install -y wget gcc g++ cmake ninja-build
# Setup build folder
WORKDIR=`pwd`
mkdir build && cd ./build
mkdir lib-tweedledum && cd ./lib-tweedledum
# Download and unpack sources
wget https://files.pythonhosted.org/packages/source/t/tweedledum/tweedledum-1.1.1.tar.gz \
https://github.com/boschmitt/tweedledum/pull/170.patch \
https://github.com/pybind/pybind11/archive/refs/tags/v2.10.4.tar.gz
tar xvf tweedledum-1.1.1.tar.gz && tar xvf v2.10.4.tar.gz
cd ./tweedledum-1.1.1
# Patch sources
patch --forward --strip=1 --input="../170.patch"
cd ./external
rm -r pybind11 && mv ../../pybind11-2.10.4 pybind11
cd ../include/tweedledum/IR
sed -i '6s/.*/#include <cstdint>/' ./Cbit.h
cd ../../..
sed -i 's/project/option/' ./pyproject.toml
# Install library
python ./setup.py build
pip install .
# Cleanup
cd ${WORKDIR} && unset WORKDIROptionally, you can check that the installation was successful by listing the installed libraries:
pip list | grep 'tweedledum'Some examples that use QASP can be found in the ./src/examples folder of this repository.
In order to run them, first install the just command runner:
snap install --edge --classic just
just -l # Show available commandsTo see a list of all the available examples, run:
just listThen, to run one, use the following command:
just run <name>