Tiled OME-Zarr

This script shows how to create a tiled single-resolution OME-Zarr dataset, make a tiles grid, and write data.

import os
from tempfile import TemporaryDirectory

import numpy as np

from iohub.ngff import open_ome_zarr

Set storage path

tmp_dir = TemporaryDirectory()
store_path = os.path.join(tmp_dir.name, "tiled.zarr")
print("Zarr store path", store_path)

Zarr store path /tmp/tmpqih40eag/tiled.zarr

Write 5D data to a new Zarr store

# define grid (rows, columns)
grid_shape = (2, 3)
# define tile shape (5D array)
tile_shape = (5, 2, 3, 32, 32)


with open_ome_zarr(
    store_path, layout="tiled", mode="a", channel_names=["DAPI", "GFP"]
) as dataset:
    dtype = np.uint16
    tiles = dataset.make_tiles(
        "tiled_raw", grid_shape=grid_shape, tile_shape=tile_shape, dtype=dtype
    )
    for row in range(grid_shape[0]):
        for column in range(grid_shape[1]):
            # each tile will be filled with different constant values
            data = np.zeros(shape=tile_shape, dtype=dtype) + row + column
            tiles.write_tile(data, row, column)

Load the dataset

with open_ome_zarr(store_path, layout="tiled", mode="r") as dataset:
    # data store structure
    dataset.print_tree()
    # get tiled image array
    tiled = dataset["tiled_raw"]
    # check grid and tile shapes
    print(tiled.tiles, tiled.tile_shape)
    # read a tile
    tile_1_2 = tiled.get_tile(1, 2)

/
 └── tiled_raw (5, 2, 3, 64, 96) uint16
(2, 3) (1, 1, 1, 32, 32)

Try viewing the images with napari-ome-zarr

Clean up

tmp_dir.cleanup()