Skip to content

Latest commit

 

History

History
122 lines (83 loc) · 3.53 KB

install-with-conda.md

File metadata and controls

122 lines (83 loc) · 3.53 KB
Raw

Installation - Conda

Getting conda

If you don’t have conda already, we suggest Miniforge, which is what we’re using.

If you already have conda through another distribution, such as Miniconda or Anaconda, that should be fine. (The environment.yml file states the channel explicitly, rather than relying on channel defaults that vary across distributions.)

Optional: Using mamba

All the following conda commands can be run with mamba instead of conda if you have mamba installed.

The exception is that, on Windows, conda activate and conda deactivate must still be used, due to limitations in mamba PowerShell integration.

For commands that create and update the environment, mamba may be faster than conda.

Creating and using the environment

To obtain and set up palgoviz:

  1. Clone the repository:

    git clone https://github.com/EliahKagan/palgoviz.git

    If you forked the repository, which you may want to do, then use your fork’s URL instead.

  2. Go into the top-level project directory:

    cd palgoviz

  3. Create the conda environment:

    conda env create

  4. Activate the environment:

    conda activate palgoviz

    You must activate the environment each time you use the project in a new shell (or after deactivating it).

  5. Create an “editable install” in the environment:

    pip install -e .

    That command only has to be run once (unless you delete and recreate the environment). It does not need to be run each time you use the project.

    This step is required to allow some modules to be run as scripts, and to be run from any location. Most of the project will work without it, but a small amount of functionality, and tests for that functionality, will fail without it.

    If you have conda-build installed, you can use conda develop . instead of pip install -e ., if you like.

Updating the environment

If you’ve just installed the project, then it is already up to date (for now). The following instructions are for later.

At this time, most project dependencies are not pinned in environment.yml, and dependencies may continue to change as work proceeds further on the project. (See “Pinning note” below.)

The following is recommended for updating the project.

  1. Activate the environment if it is not already active:

    conda activate palgoviz

  2. Update packages currently installed in the environment:

    conda update --all

  3. Make sure you are in the top-level directory. That is the directory that contains the project-wide README and environment.yml files.

  4. Install any new project dependencies, and remove any old dependencies that are known to no longer be needed:

    conda env update --prune

Pinning note

We are not currently pinning specific package versions, nor listing indirect dependencies, in environment.yml.

If you want dependencies pinned to specific known-working versions, you can install with poetry instead of with conda.