Installation

Creating the DP3 / WSClean singularity image

The pipeline is capable of running DP3 and WSClean inside a singularity container. If you wish to use that feature, a requirement is thus to build a singularity image file that allows to run them both.

Please follow the instructions on the Docker and singularity images page.

Installing the pipeline

A highly recommended first step is to create an isolated Python environment in which to install the pipeline’s Python module. There are several ways of doing this, so feel free to choose your favourite. Here we will use conda because that is often the environment manager of choice when working on supercomputing facilities.

The pipeline requires Python version 3.9 or higher. Below we create a new conda environment called selfcal, activate it, then inside it we install the official package manager of SKAO: poetry.

conda create --name selfcal python=3.9
conda activate selfcal
conda install poetry

Then, navigate to the base directory where you wish to clone the repository and run:

git clone --recurse-submodules https://gitlab.com/ska-telescope/sdp/science-pipeline-workflows/ska-sdp-wflow-mid-selfcal

Note

--recurse-submodules is required because the SKAO Makefiles are provided as a git submodule which must be pulled explicitly.

Lastly, install the module:

cd ska-sdp-wflow-mid-selfcal/
poetry install

The poetry install command does the equivalent of pip’s editable install, meaning that you may freely edit the code and see the changes in action next time the pipeline code is run.

Once the installation is complete, the pipeline command-line app should be accessible from anywhere when inside your newly created Python environment. Type the following commands to check both direction-independent and direction-dependent pipelines were successfully installed:

mid-selfcal-di --help
mid-selfcal-dd --help