This script takes paths to raw/label n5 datasets as an input, and outputs .zarr container with the correct cellmap schema:

n5 / zarr container

• recon_{number}
  • em
    • fibsem-uint8
    • fibsem-uint16
  • labels
    • inference (predictions and segmentations)
    • groundtruth (training crops)
    • mask

Installation:

1. cd PATH_TO_POETRY_PROJECT_DIRECTORY/
2. poetry install

Example(for lsf cluster):

bsub -n 15 -J n5convert -o path_to_output_log_file 'poetry run python src/to_cellmap_zarr.py --num_workers=300 --cluster=lsf --src=path_to_raw_data_n5_group(array) --inf=path_to_inference --dest=path_to_the_output_zarr_container';