Required packages are specified in `environment.yml`.
An environment can be created with the required packages using the conda
package manager:
We strongly recommend installing within a python environment as there are many dependencies that you may not want polluting your regular python environment.
An environment can be created with using the conda package manager:
Alternatively, you can install the package locally as described below.
$ conda create --name <env>
$ conda activate <env>
## Local installation
Which you can deactivate with:
Run the following from outside of the pipline directory to install an editable
version of the pipeline on your local machine.
$ conda deactivate
It is still recommended to install in a conda or virtualenv environments.
Or using virtualenv:
```
conda create -n <env> python=3.6
conda activate <env>
pip install -e pipeline-core
```
$ python -m virtualenv /path/to/venv/
$ source /path/to/venv/bin/activate
This will download all of your packages under `/path/to/venv` and then activate it.
Deactivate using
$ deactivate
## Pipeline installation
Once you have created your local environment, run:
$ cd pipeline-core
$ pip install -e ./
You will be asked to put in your GitLab credentials for a couple of the packages installed as dependencies.
The `-e` option will install in 'editable' mode: all of the changes that are made to the code in the repository will immediately be reflected in the installation.
This is very useful to keep up with any changes and branching that we make.
That way, if you want to keep up with the most recent updates, just run:
$ git pull
If you would rather be more in charge of which updates you install and which you don't, remove `-e` from your installation command.
In this case (or if you run into a dependency error) in order to update your installed version you will have to run:
