Issues intalling with pip/conda

[stevenstetzler]

I have run into issues installing the requirements using conda/pip. I initially tried installing with pip using requirements.txt and dev-requirements.txt per the README:

conda create -n precovery python=3.10
conda activate precovery
# python3.10 -m venv virtualenv3.10 # ignore since I installed python3.10 with conda
# source virtualenv3.10/bin/activate # ignore since I made a conda env
pip install pip --upgrade
pip install -r requirements.txt -r dev-requirements.txt
pre-commit install

gave errors saying that llvmlite==0.36.0 and numba==0.53.1 were not available, making the fixed versions impossible to install.

I think installed using the requirements.in and dev-requirements.in files, where only numpy and numba are versioned and continued to receive an error since numba==0.53.1 is defined in requirements.in. I replaced numba==0.53.1 with numba==0.55.1 which led to a solve. Finally, pip installed numpy version 1.22 but numba threw errors on execution claiming in needs numpy<=1.21. I solved this with:

pip install numpy==1.21

And with that the precovery code ran.

Additionally, I needed to install pyoorb:

conda install openorb

I can make updates to the README and a pull request at a later time.

[moeyensj]

We relaxed the version requirements in PR #38, which has the side effect of fixing this issue. If problems persist please feel completely free to reopen here!