Converts electrophysiology, photometry, and behavioral data for the hex maze task used by the Berke Lab at UCSF to NWB format for sharing and analysis.
git clone https://github.com/calderast/jdb_to_nwb.git
cd jdb_to_nwb
pip install -e .-
Open
metadata_example.yamlin a text editor. Update the paths to point to your data and update the metadata for your experiment. -
Run the conversion to generate an NWB file (replace
output_dirwith your desired output directory). The nwb file will be automatically named based on the animal name and date (i.e.rat_date.nwb):
jdb_to_nwb metadata_example.yaml output_dir- Sub-directories for associated figures and conversion log files will be created alongside the nwb file in
output_dir. Check that there are no errors in the error log file and that all figures look as expected.
The large test data files are stored in a shared UCSF Box account. To get access to the test data, please contact the repo maintainers.
Create a new file called .env in the root directory of the repository and add your Box credentials:
BOX_USERNAME=<your_box_username>
BOX_PASSWORD=<your_box_password>Or set the environment variables in your shell:
export BOX_USERNAME=<your_box_username>
export BOX_PASSWORD=<your_box_password>Then run the download script:
python tests/download_test_data.pyYou can pass the --overwrite flag to overwrite existing files:
python tests/download_test_data.py --overwriteNotes:
- Run
python tests/test_data/create_raw_ephys_test_data.pyto re-create the test data forraw_ephys. - Run
python tests/test_data/create_processed_ephys_test_data.pyto re-create the test data forprocessed_ephys. tests/test_data/processed_ephys/impedance.csvwas manually created for testing purposes.tests/test_data/processed_ephys/geom.csvwas manually created for testing purposes.- Some files (
settings.xml,structure.oebin) nested withintests/test_data/raw_ephys/2022-07-25_15-30-00were manually created for testing purposes.
The GitHub Actions workflow (.github/workflows/test_package_build.yml) will automatically download the test data and run the tests.
Versioning is handled automatically using hatch-vcs using the latest tag in the git history as the version number. To make a new release, simply tag the current commit and push to the repository. Use semantic versioning to set the version number. Create a GitHub release using the tag.