np-aind-metadata

Usage

conda create -n np_aind_metadata python>=3.10
conda activate np_aind_metadata
pip install np_aind_metadata

On-prem usage

If on Allen Institute premises, additional features are available via np-config. To use on-prem features install np-config.

pip install np-config

Python

On-prem

Copies the latest rig.json from storage to a local path.

>>> from pathlib import Path
>>> import datetime
>>> from np_aind_metadata import rigs

>>> rigs.copy_rig("NP3", output_path=Path("rig.json"))
PosixPath('rig.json')

Off-prem

Updates a rig json and store it at output path.

>>> from pathlib import Path
>>> from np_aind_metadata.update import update_rig
>>> update_rig(
...     Path(".", "examples", "rig.json"),
...     open_ephys_settings_sources=[
...         Path(
...             ".",
...             "examples",
...             "example-session-directory",
...             "settings.xml"
...         ),
...     ],
...     output_path=Path("rig.json"),
... )
PosixPath('rig.json')

Convenience functions

Direct support for common use cases.

np-codeocean

Parses a dynamic routing session directory for a session.json, adds it's corresponding rig.json, and updates session.json to reflect the rig.json it is associated with. If not onprem, rig_storage_dir will need to be supplied.

>>> from pathlib import Path
>>> from np_aind_metadata.integrations import dynamic_routing_task
>>> session_dir = Path("examples") / "example-session-directory"
>>> dynamic_routing_task.add_rig_to_session_dir(session_dir, datetime.date(2024, 4, 1))

Model storage

Models are stored locally on disk. To initialize model storage.

Initialize storage for all supported rigs.

np-aind-metadata init-rig-storage "/directory/to/store/rigs"

Initialize storage for a specific rig.

np-aind-metadata init-rig-storage "/directory/to/store/rigs" "NP3" --date 2022/02/07

Update a stored rig.

np-aind-metadata update-rig "/directory/to/store/rigs" "NP3" --date 2022/02/07

Local development

Testing

Testing intended for cloned project from source control.

Unit tests

pdm run pytest

Storage tests

pdm run pytest-storage

On-prem tests

Requires user to likely be on prem with np group dependencies installed.

Install np group dependencies

pdm install -G np

Run tests

pdm run pytest-onprem

All tests

Requires dependencies from onprem.

pdm run pytest-full

Development

See instructions in https://github.com/AllenInstitute/np-aind-metadata/CONTRIBUTING.md and the original template: https://github.com/AllenInstitute/copier-pdm-npc/blob/main/README.md