installation.rst

Installation

Note

If you are on the BIH cluster, first read :ref:`quickstart` as this also explains the temporary directory.

Prerequisites

The CUBI pipeline requires Python >=3.12 (e.g., from a Miniconda3 installation).

More recent versions also work but other requirements as Snakemake might make it depend on a more recent Python version.

For cluster execution, you need a Snakemake profile available.

Installing a Release

This is the recommended way if you just want to use the pipeline, simply read :ref:`quickstart`.

Installing as a Developer

It is highly recommended to have a Miniconda installation for the development as this allows for easily resetting everything. You can of course clone the code anywhere you like.

$ mkdir -p ~/Development/pipeline_dev
$ cd ~/Development/pipeline_dev
$ git clone [email protected]:bihealth/snappy-pipeline.git
$ cd snappy_pipeline
$ conda env create -n snappy_dev --file environment.yaml
$ conda activate snappy_dev
$ pip install -e ".[all]"

Installing pre-commit-hooks

To make it easier to follow the coding style, we use pre-commit hooks. These hooks will run the style checks before you commit your changes and will automatically fix some issues.

First, install the pre-commit package (if not already installed, part of the optional dependency group dev):

$ conda install pre-commit  # or pip install pre-commit

Then, install the pre-commit hooks:

$ pre-commit install

The next time you commit changes, the pre-commit hooks will run automatically.

Running the Tests

To run the tests, simply invoke pytest (part of the optional dependency group test):

$ cd ~/Development/pipeline_dev
$ pytest

Running the Style Checks

$ cd ~/Development/pipeline_dev
$ make lint

Developer Documentation

Make sure to also read the "Pipeline Developer Docs" section, starting with :ref:`dev_intro`.