This simple capsule is designed to create an NWB file with basic subject and session information.

The data/ folder must include the session dataset and optionally the following JSON files:

• subject.json: Subject information following the aind-data-schema specification
• data_description.json: session information following the aind-data-schema specification

The data_description.json is only used to access the name field, which is used as session name. An minimal example data_description.json file:

{
    "name": "my-awesome-session-name"
}

If these files are missing, a mock NWB file is created.

If the data/ folder includes an NWB file, this is copied to the results folder.

The code/run script takes 2 arguments:

  --backend {hdf5,zarr}
                        NWB backend. It can be either 'hdf5' or 'zarr'.
  --asset-name ASSET_NAME
                        Path to the data asset of the session. When provided, the metadata are fetched from the AIND metadata database. If None, and the attached data asset is used to fetch relevant metadata.

The --asset-name input is only used at AIND to fetch metadata from the database.

The output of this capsule is a NWB file in the results/ folder named: {session_name}.nwb

In case the data/ folder contains a single NWB file, this is copied to the results with the same name.