NeuRosetta

Create a usable environment (quick version while I'm re-building)

If you have a CUDA enable GPU, make sure CUDA is usable!

• make sure you have miniconda available. install instructions can be found here
• I've used mamba for faster install, so run conda install -n base -c conda-forge mamba -y if you do NOT have mamba
• Build your environment, i've called this one nr:

mamba create -n nr python=3.11 -y \
  numpy pandas scipy trimesh vedo gudhi ipython \
  -c conda-forge \
  graph-tool

This creates an environment with numpy, pandas, scipy, trimesh, vedo, gudhi, and graph-tool.

• Activate the environment:

conda activate nr

• install Jax with GPU using pip:

python3 -m pip install --upgrade
python3 -m pip install "jax[cuda13]"

• ALTERNATIVE Jax without a GPU:

python3 -m pip install --upgrade
python3 -m pip install jax

Running tests

from root folder:

python3 -m pytest
``