IO (sphero_vem.io)#

Zarr and TIFF I/O with OME-NGFF metadata tracking.

Module containing input/output functions

sphero_vem.io.write_image(fname, image, compressed=False, **kwargs)[source]#

Save a NumPy array as a TIFF file, optionally with zlib compression.

Parameters:

  • fname (Path) – Destination file path.

  • image (numpy.ndarray) – Image array to save.

  • compressed (bool, optional) – If True, apply zlib compression (level 6) with tiling. Extra keyword arguments are ignored when compression is enabled. Default is False.

  • **kwargs – Additional keyword arguments forwarded to tifffile.imwrite when compressed is False.

Return type:

None

sphero_vem.io.stack_to_zarr(stack_dir, root_path, spacing, chunk_size=(1, 1024, 1024), verbose=True)[source]#

Convert a tiff stack to a ZYX zarr archive.

The stack will be saved under root/images/spacing_dir

Parameters:

  • stack_path (Path) – Path to the tiff stack. This should be a directory with single tif slices that will be concatened along the Z axis.

  • dest_path (Path) – Path the root of the destination zarr store.

  • spacing (tuple[int, int, int] | None) – ZYX spacing of the dataset in nanometers. If None, attempt to read the spacing from metadata (Currently not implemented).

  • chunk_size (tuple[int, int, int] | None) – ZYX chunk size of the zarr array. If None, use (1, H, W).

  • verbose (bool) – Enable verbose output.

  • stack_dir (Path)

  • root_path (Path)

Raises:

NotImplementedError – When passing None to spacing.

Return type:

None

sphero_vem.io.write_zarr(root, array, dst_path, src_zarr=None, spacing=None, dtype=None, shape=None, processing=None, inputs=None, zarr_chunks=None)[source]#

Write a numpy or dask array to zarr with metadata.

Parameters:

  • root (zarr.Group | Path | str) – Root zarr group, or path to one.

  • array (np.ndarray | da.Array) – Array to save. Axis order should be (CZ)YX.

  • dst_path (str) – Path under root where to save the array.

  • src_zarr (zarr.Array | None) – Source zarr array. Used to read spacing, chunks, previous processing, and inputs if not explicitly provided. Default is None.

  • spacing (tuple[int | float, ...] | None) – Spacing of the array. If None, reads from src_zarr. Default is None.

  • dtype (Any | None) – dtype to cast to when saving. If None, uses array dtype. Default is None.

  • shape (tuple[int, ...] | None) – Zarr array shape. If None, uses array shape. Default is None.

  • processing (ProcessingStep | list[ProcessingStep] | list[dict] | dict | None) – Processing steps to append to src_zarr processing history. Default is None.

  • inputs (list[str] | None) – Paths to input arrays. If None, uses src_zarr path. Default is None.

  • zarr_chunks (tuple[int, ...] | None) – Chunk shape. If None, reads from src_zarr. Default is None.

Raises:

  • ValueError – If zarr_chunks is None and src_zarr is None.

  • ValueError – If spacing is None and src_zarr is None or has no spacing attribute.

  • TypeError – If array is neither a numpy nor a dask array.

Return type:

None