Environment Setup

This directory provides scripts to automate the creation of a conda environment for reproducible runs of the Tutorial notebooks in both python and R.

To create the conda environment, simply run the bash script using the following command:

bash -i setup_env.sh

This will create the environment with both Python and R dependencies, but it does not enable a GPU.

The following optional flags are available as well:

• -c | --conda: use the conda command to generate the environment, otherwise defaults to mamba
• -k | --kernel: install jupyter kernels to run notebooks in the environment
• -g | --gpu: install necessary dependencies to run a cuda-enabled GPU
• -p | --python-only: sets up only the python-associated dependencies. Use this if you do not plan using R.
• -n | --name <ENV_NAME>: specify the environment name, otherwise defaults to "ccc_protocols"

In case of issues, .yml files are commented to include exact versions and builds used.

Examples of other setups:

Install Python and R dependencies enabling a GPU and kernel for jupyter notebooks (using conda):

bash -i setup_env.sh --conda --gpu --kernel

Install only Python dependencies and enable kernel for jupyter notebooks (using conda):

bash -i setup_env.sh --conda --python-only --kernel

Install only Python dependencies, enable a GPU and kernel for jupyter notebooks (using conda):

bash -i setup_env.sh --conda --python-only --gpu --kernel