This repository will review the steps to build a Singularity container from a definition file and implement it into scripts. Furthermore, this repository also covers developing and implementing a configuration file into scripts. Singularity containers and configuration files aim to make coding workflows more efficient and portable. Integrating these features into scripts will be a significant time-saver for research and data analysis work.
These tutorials are Jupyter Notebooks; some features, such as internal links, may not render in GitHub correctly. Please copy and paste the notebook URL into nbviewer for full compatibility.
Recently the creator of anaconda has started demanding payment from academic research group for the continued use of the non-free repositories. To work around this we have moved to miniforge3 instead of miniconda3. Miniforge3 is a srop in replacement for miniconda3, but avoids the legal issues.
In the Creating-a-Singularity-Container-for-a-Conda-Environment instructions it mentions the use of miniconda3. Those instructions will be corrected at a different time, for now what you should know is instead of using the provided Singularity definition file instead use the template available at /data/projects/containers/miniconda/Singularity.template.def.
Following the rough instuctions found above do:
cd /path/to/my/project/
cp /data/projects/containers/miniconda/Singularity.template.def ./Singularity.def
# Optional, select a different base OS by uncommenting lines 3 or 4 and commenting out line 5.
# You likely want to use the default Rocky Linux 9
conda list -p /path/to/my/project/environment --explict > requirements.txt
apptainer build MyContainer.sif Singularity.def