Pybids databases let you read BIDSLayouts in the blink of an eye, and are practically mandatory when iteratively developing workflows using tools like snakebids. This library implements a command-line solution to generating these layouts.
By default, the database is generated on /tmp, which is assumed to mounted to a local or in-memory filesystem. This is important for datasets stored on NAS, as writing SQL to these systems takes a very, very, very long time. The library also enforces a special integration with AllianceCan's compute network: if SLURM commands like sbatch or srun are available, the app checks if the $SLURM_TMPDIR variable is defined to make sure a compute node is being used. If the variable is found, it will be used to generate the database.
In any case, after generation, the database is moved to the .pybids directory in the top level of your dataset, or in another directory of your choosing.
Usage: pybidsdb [options] [path]
Arguments:
path: Path of the dataset (defaults to current directory)
--force: Force run on the login node (by default only runs on compute nodes)
--validate: Run bids validation on the dataset
--derivatives: Include derivatives found in the `dataset/derivatives` folder
--folder: Name of the database folder (e.g. `dataset/<folder>`)
Install with pipx:
pipx install git+https://github.com/pvandyken/pybidsdb.git