Skip to content

Exercise 2: Convert Images Locally

Activate the conda environment:

source /scratch/miniforge3/etc/profile.d/conda.sh      # source the conda binary
conda activate eubizarr                                # activate the eubizarr environment
  • You could list all the available conda enviroments using conda env list.
  • Always check you are using the correct conda environment.

Create a results folder:

mkdir ~/results_local
  • The ~ expand to your own home folder, e.g. /home/YOUSUSERNAME/
  • On BAND, you only have write permission to your home folder and /scratch
  • You can create any folder under ~ or /scratch to store results.

Convert several single images

eubi to_zarr /scratch/common/data/small_dataset/ ~/results_local
  • These steps convert all images in the small_dataset folder into Zarr format and save the results in ~/results_local.
  • The conversion only uses 1 node.
  • The conversion runs on the same node where your BAND desktop is running.

Check the results

List the contents of the output folder:

ls -lah ~/results_local

ome_zarr info ~/results_local/ftsz3.zarr

Expect Output

queue

Convert one multi-scene image

  1. Configure eubi multi-scene conversion (should persist and not be necessary for the course)
eubi configure readers --scene_index all

  1. Open /scratch/common/data/medium_dataset/19_07_19_Lennard.lif in Fiji to observe that it is multi-scenece container queue

  2. Convert to OME-Zarr

Create output folder 

mkdir ~/medium_data_out

Run the conversion

/scratch/common/data/medium_dataset/19_07_19_Lennard.lif ~/medium_data_out --max_concurrent_scenes 10

queue