conda-env-tracker makes it simple to keep your software environment up to date. Manage python, R, and conda packages (of any language) for multiple projects with ease.
- python >= 3.6
- Anaconda/Miniconda install
- conda >= 4.5
If you are using an old version of conda we recommend updating:
$ conda update -n base conda
$ conda init
Currently you can only install directly from GitHub by cloning and running pip install conda-env-tracker/. We recommend installing in your base conda environment.
- Initialize the conda-env-tracker command line tool. Only needs to be done once, right after installing. Using
--autorunscet autowhich will automatically check if you need to be synced with the conda-env-tracker environment in your current git repo.
$ cet init --auto
- Clone or navigate to a git repo:
$ git clone git@github.com:Name/my-repo.git # If necessary
$ cd my-repo
conda-env-tracker sync changes to 'my-env' environment ([y]/n/stop asking)? y
...
Activate the 'my-env' environment ([y]/n/stop asking)? y
(my-env) $
If there is a conda-env-tracker environment in the repo, then you will be asked to sync the environment and activate it. Otherwise learn how to create an environment in the next section.
Create a new environment from inside a git repo:
$ cet create --sync --name my-env python=3.7 pandas
...
Activate the 'my-env' environment ([y]/n/stop asking)?
sync and commit the changes:
$ git add .cet/*
$ git commit -m "Add cet"
Activate an environment (conda activate my-env if necessary) and run
(my-env) $ cet install jupyter
or (only recommended for packages not available via conda)
(my-env) $ cet pip install arcgis
or (similarly, only recommended for packages not available via conda)
(my-env) $ cet r install testthat --command 'install.packages("testthat")'
If you are in the git repo associated with this environment conda-env-tracker will automatically suggest a sync, but you can force one using
(my-env) $ cet sync