Configuration¶

Click any command card below to expand its full parameter reference.

Info

Note that all cards below are auto-generated from EuBIBridge methods and Pydantic config models. Types, defaults, and valid ranges are extracted directly from the source.

configure cluster — Update cluster / concurrency parameters. Omitted arguments keep their current values

Usage:

eubi configure cluster [OPTIONS]

Optional arguments

--max_workers

Type: int

Default: default

Number of parallel file-level worker processes (default 4).

# Set the default number of parallel worker processes
eubi configure cluster --max_workers 8

--queue_size

Type: int

Default: default

Internal write-queue depth per worker (default 4).

# Allow each worker to buffer more write tasks
eubi configure cluster --queue_size 8

--region_size_mb

Type: int

Default: default

Region size in MB for spatial partitioning (default 256).

# Process larger spatial regions per task (reduces overhead on large arrays)
eubi configure cluster --region_size_mb 512

--memory_per_worker

Type: str

Default: default

Memory limit accepted by SLURM / LocalCluster, e.g. '8GB' (default '1GB').

# Allocate 8 GB per worker (relevant for SLURM / LocalCluster)
eubi configure cluster --memory_per_worker 8GB

--max_concurrency

Type: int

Default: default

TensorStore write concurrency per worker (default 4).

# Allow TensorStore to issue 8 write operations concurrently per worker
eubi configure cluster --max_concurrency 8

--max_concurrent_scenes

Type: int

Default: default

Parallel scenes within one file (default 1). Increase only when writing a multi-scene file to a large store.

# Convert 2 scenes within a single file simultaneously
eubi configure cluster --max_concurrent_scenes 2

--max_concurrent_downscale_layers

Type: int

Default: default

Number of downscaled pyramid layers written concurrently per scene (default 3).

eubi to_zarr /data/input /data/output --max_concurrent_downscale_layers 2

--on_local_cluster

Type: boolean flag

Default: default

Valid values: --on_local_cluster to enable · --on_local_cluster False to disable

Use a Dask LocalCluster backend (default False).

# Use a Dask LocalCluster as the scheduler
eubi configure cluster --on_local_cluster

--on_slurm

Type: boolean flag

Default: default

Valid values: --on_slurm to enable · --on_slurm False to disable

Submit jobs to a SLURM cluster (default False).

# Submit workers to SLURM and set partition + account
eubi configure cluster --on_slurm --slurm_account myaccount --slurm_partition gpu

--use_threading

Type: boolean flag

Default: default

Valid values: --use_threading to enable · --use_threading False to disable

Use ThreadPool instead of ProcessPool (default False).

# Switch from ProcessPool to ThreadPool (useful when GIL is not a bottleneck)
eubi configure cluster --use_threading

--tensorstore_data_copy_concurrency

Type: int

Default: default

TensorStore internal copy threads (default 4).

# Allow TensorStore to copy data with 8 internal threads
eubi configure cluster --tensorstore_data_copy_concurrency 8

--max_retries

Type: int

Default: default

Retries on broken worker process (default 10).

# Retry failed workers up to 5 times before aborting
eubi configure cluster --max_retries 5

--bf_read_concurrency

Type: int

Default: default

Dask thread count for parallel bfio tile reads (default 4). None lets dask choose (cpu_count).

--bf_tile_size_mb

Type: float

Default: default

Tile size budget in MB for bfio tiled reading (default 512).

--jvm_memory

Type: str

Default: default

Maximum JVM heap for Bio-Formats, e.g. '8GB', '4GB'. Accepts 'NGB' / 'NMB' (like memory_per_worker) and normalises internally to JVM format ('Ng' / 'Nm'). Default '2g'.

--slurm_time

Type: str

Default: default

SLURM job wall-clock limit, e.g. '24:00:00' (used only when on_slurm=True).

# Set SLURM wall-clock limit to 48 hours
eubi configure cluster --on_slurm --slurm_time 48:00:00

--slurm_account

Type: str

Default: default

SLURM account/allocation to charge (default None).

# Charge SLURM jobs to 'myproject'
eubi configure cluster --on_slurm --slurm_account myproject

--slurm_partition

Type: str

Default: default

SLURM partition/queue to submit to (default None).

# Submit to the 'gpu' partition
eubi configure cluster --on_slurm --slurm_partition gpu

--slurm_worker_timeout

Type: int

Default: default

Seconds to wait for SLURM workers to start (default 300).

# Wait up to 10 minutes for SLURM workers to start
eubi configure cluster --on_slurm --slurm_worker_timeout 600

--slurm_sif_path

Type: str

Default: default

Path to an Apptainer/Singularity SIF image to run workers in (default None).

eubi to_zarr /data/input /data/output --on_slurm --slurm_sif_path /apps/eubi.sif

configure conversion — Update Zarr conversion parameters. Omitted arguments keep their current values

Usage:

eubi configure conversion [OPTIONS]

Optional arguments

--ome_zarr_version

Type: str

Default: default

OME-Zarr (NGFF) version to write, e.g. '0.4' or '0.5'. This is the preferred control and supersedes zarr_format (the zarr container format is derived from it).

# Default to OME-Zarr 0.5 (Zarr v3, enables sharding)
eubi configure conversion --ome_zarr_version 0.5

# Default to OME-Zarr 0.4 (Zarr v2)
eubi configure conversion --ome_zarr_version 0.4

--zarr_format

Type: int

Default: default

DEPRECATED zarr container version — 2 (default) or 3. Used only when ome_zarr_version is unset/unrecognised; prefer ome_zarr_version.

# Deprecated — prefer 'ome_zarr_version'. Switch to Zarr v3 (supports sharding)
eubi configure conversion --zarr_format 3

--skip_dask

Type: boolean flag

Default: default

Valid values: --skip_dask to enable · --skip_dask False to disable

Read TIFF files via zarr's native aszarr backend instead of dask. Faster for large TIFFs; ignored for non-TIFF formats.

# Use zarr's native TIFF backend instead of dask (faster for large TIFFs)
eubi configure conversion --skip_dask

--auto_chunk

Type: boolean flag

Default: default

Valid values: --auto_chunk to enable · --auto_chunk False to disable

Auto-compute chunk shape from array size (default True).

# Disable auto-chunking and set chunk sizes manually
eubi configure conversion --auto_chunk False --z_chunk 64 --y_chunk 256 --x_chunk 256

--target_chunk_mb

Type: float

Default: default

Target chunk size in MB when auto_chunk is True.

# Target 4 MB uncompressed chunks when auto_chunk is on
eubi configure conversion --target_chunk_mb 4.0

--time_chunk

Type: int

Default: default

Manual per-axis chunk sizes (used when auto_chunk is False).

# Process one time point per chunk
eubi configure conversion --auto_chunk False --time_chunk 1

--channel_chunk

Type: int

Default: default

Chunk size along the channel axis. Applies only when auto_chunk=False.

# Process one channel per chunk
eubi configure conversion --auto_chunk False --channel_chunk 1

--z_chunk

Type: int

Default: default

Chunk size along the z axis. Applies only when auto_chunk=False.

# Set manual z chunk size (requires auto_chunk False)
eubi configure conversion --auto_chunk False --z_chunk 64

--y_chunk

Type: int

Default: default

Chunk size along the y axis. Applies only when auto_chunk=False.

# Set manual y chunk size (requires auto_chunk False)
eubi configure conversion --auto_chunk False --y_chunk 256

--x_chunk

Type: int

Default: default

Chunk size along the x axis. Applies only when auto_chunk=False.

# Set manual x chunk size (requires auto_chunk False)
eubi configure conversion --auto_chunk False --x_chunk 256

--time_shard_coef

Type: int

Default: default

Shard-to-chunk multipliers for Zarr v3 sharding (default 3 on spatial axes).

# Shard = 1 × time chunk (no sharding on time axis)
eubi configure conversion --ome_zarr_version 0.5 --time_shard_coef 1

--channel_shard_coef

Type: int

Default: default

Shard size = chunk × coef for the channel axis. Zarr v3 only — ignored for v2.

# Shard = 1 × channel chunk
eubi configure conversion --ome_zarr_version 0.5 --channel_shard_coef 1

--z_shard_coef

Type: int

Default: default

Shard size = chunk × coef for the z axis. Zarr v3 only — ignored for v2.

# Shard = 5 × z chunk on the z axis
eubi configure conversion --ome_zarr_version 0.5 --z_shard_coef 5

--y_shard_coef

Type: int

Default: default

Shard size = chunk × coef for the y axis. Zarr v3 only — ignored for v2.

# Shard = 5 × y chunk on the y axis
eubi configure conversion --ome_zarr_version 0.5 --y_shard_coef 5

--x_shard_coef

Type: int

Default: default

Shard size = chunk × coef for the x axis. Zarr v3 only — ignored for v2.

# Shard = 5 × x chunk on the x axis
eubi configure conversion --ome_zarr_version 0.5 --x_shard_coef 5

--time_range

Type: int

Default: default

Crop ranges as "start,stop" strings applied before writing.

# Keep only time frames 0–9 (start inclusive, stop exclusive)
eubi configure conversion --time_range "0,10"

--channel_range

Type: int

Default: default

Crop the channel axis. Same format as time_range, e.g. "0,2" keeps the first two channels.

# Keep only the first two channels
eubi configure conversion --channel_range "0,2"

--z_range

Type: int

Default: default

Crop the z axis. Same format as time_range, e.g. "5,50" keeps z-slices 5–49.

# Keep z-slices 5 through 49
eubi configure conversion --z_range "5,50"

--y_range

Type: int

Default: default

Crop the y axis. Same format as time_range.

# Crop y to the first 512 pixels
eubi configure conversion --y_range "0,512"

--x_range

Type: int

Default: default

Crop the x axis. Same format as time_range.

# Crop x to the first 512 pixels
eubi configure conversion --x_range "0,512"

--compressor

Type: str

Default: default

Compression codec — 'blosc' (default), 'gzip', 'zstd', 'bz2', or 'none'.

# Use zstd compression (good balance of speed and ratio)
eubi configure conversion --compressor zstd

# Disable compression entirely
eubi configure conversion --compressor none

--compressor_params

Type: dict

Default: default

Dict of codec parameters, e.g. {'cname': 'lz4', 'clevel': 5}.

# Use blosc with lz4 codec at compression level 9
eubi configure conversion --compressor blosc --compressor_params '{"cname": "lz4", "clevel": 9}'

--overwrite

Type: boolean flag

Default: default

Valid values: --overwrite to enable · --overwrite False to disable

Overwrite existing output zarr (default False).

# Allow overwriting an existing OME-Zarr at the output path
eubi configure conversion --overwrite

--override_channel_names

Type: boolean flag

Default: default

Valid values: --override_channel_names to enable · --override_channel_names False to disable

Replace output channel labels with the channel_tag values. For aggregative conversions with a tuple channel_tag only.

# Replace channel labels with the channel_tag values from filenames
eubi configure conversion --override_channel_names

--channel_intensity_limits

Type: Literal['from_dtype', 'from_array', 'auto']

Default: default

How to set OMERO window limits — 'from_dtype' (default) uses dtype min/max, 'from_array' computes per-channel min/max from pixel data, 'auto' lets the viewer decide.

# Compute display window limits from actual pixel data
eubi configure conversion --channel_intensity_limits from_array

# Let the viewer compute limits automatically
eubi configure conversion --channel_intensity_limits auto

--metadata_reader

Type: str

Default: default

Metadata backend — 'bfio' (default) or 'bioformats'.

# Use the Java BioFormats backend for metadata (more formats supported)
eubi configure conversion --metadata_reader bioformats

--save_omexml

Type: boolean flag

Default: default

Valid values: --save_omexml to enable · --save_omexml False to disable

Write a companion OME-XML file alongside the zarr (default True).

# Disable OME-XML sidecar file generation
eubi configure conversion --save_omexml False

--squeeze

Type: boolean flag

Default: default

Valid values: --squeeze to enable · --squeeze False to disable

Remove singleton dimensions before writing (default True).

# Keep length-1 dimensions instead of removing them
eubi configure conversion --squeeze False

--dtype

Type: str

Default: default

Output dtype — 'auto' keeps the source dtype, or any NumPy dtype string such as 'uint16'.

# Cast pixel values to uint8 on write (reduces file size)
eubi configure conversion --dtype uint8

# Keep the source dtype
eubi configure conversion --dtype auto

--verbose

Type: boolean flag

Default: default

Valid values: --verbose to enable · --verbose False to disable

Log verbose progress output (default False).

# Enable verbose per-chunk progress logging
eubi configure conversion --verbose

configure downscale — Update downscale parameters. Omitted arguments keep their current values

Usage:

eubi configure downscale [OPTIONS]

Optional arguments

--n_layers

Type: int

Default: default

Number of downscale pyramid levels to generate. None (default) auto-derives the count from min_dimension_size.

# Generate 5 pyramid levels
eubi configure downscale --n_layers 5

--min_dimension_size

Type: int

Default: default

Stop adding levels when the smallest spatial dimension falls below this pixel count (default 64).

# Stop downscaling when the smallest spatial dimension is below 32 px
eubi configure downscale --min_dimension_size 32

--time_scale_factor

Type: int

Default: default

Downscale factor along the time axis (default 1, i.e. no downscaling).

# Disable time-axis downscaling (value 1 = no downscaling)
eubi configure downscale --time_scale_factor 1

--channel_scale_factor

Type: int

Default: default

Downscale factor along the channel axis (default 1).

# Disable channel-axis downscaling
eubi configure downscale --channel_scale_factor 1

--z_scale_factor

Type: int

Default: default

Downscale factor along the z axis (default 2).

# Isotropic downscaling on all spatial axes
eubi configure downscale --z_scale_factor 2 --y_scale_factor 2 --x_scale_factor 2

# Anisotropic — skip z downscaling for thick sections
eubi configure downscale --z_scale_factor 1 --y_scale_factor 2 --x_scale_factor 2

--y_scale_factor

Type: int

Default: default

Downscale factor along the y axis (default 2).

eubi configure downscale --y_scale_factor 2

--x_scale_factor

Type: int

Default: default

Downscale factor along the x axis (default 2).

eubi configure downscale --x_scale_factor 2

--downscale_method

Type: str

Default: default

Pixel aggregation method for downscaling — one of 'simple', 'mean', 'median', 'min', 'max', 'mode' (default 'simple').

# Use mean aggregation (recommended for fluorescence images)
eubi configure downscale --downscale_method mean

# Use nearest-neighbour striding (fastest, recommended for label images)
eubi configure downscale --downscale_method simple

--keep_existing_resolutions

Type: boolean flag

Default: default

Valid values: --keep_existing_resolutions to enable · --keep_existing_resolutions False to disable

If the input format already carries its own multiscale pyramid (e.g. .ims, .zarr), write each existing resolution level straight to the output OME-Zarr instead of rebuilding the pyramid with the above scale factors (default False).

# Copy a source pyramid (.ims / .zarr) verbatim instead of rebuilding it
eubi configure downscale --keep_existing_resolutions

--apply_smart_downscaling

Type: boolean flag

Default: default

Valid values: --apply_smart_downscaling to enable · --apply_smart_downscaling False to disable

Make the first pyramid level use automatically-computed factors that drive the voxels toward isotropy (default False).

# Pick per-axis downscale factors automatically from voxel anisotropy
eubi configure downscale --apply_smart_downscaling

--time_smart_scale_factor

Type: int

Default: default

Per-axis override for smart downscaling on the time axis (default None = auto).

eubi to_zarr /data/input /data/output --apply_smart_downscaling --time_smart_scale_factor 1

--z_smart_scale_factor

Type: int

Default: default

Per-axis smart-downscaling override for z.

eubi to_zarr /data/input /data/output --apply_smart_downscaling --z_smart_scale_factor 1

--y_smart_scale_factor

Type: int

Default: default

Per-axis smart-downscaling override for y.

eubi to_zarr /data/input /data/output --apply_smart_downscaling --y_smart_scale_factor 2

--x_smart_scale_factor

Type: int

Default: default

Per-axis smart-downscaling override for x.

eubi to_zarr /data/input /data/output --apply_smart_downscaling --x_smart_scale_factor 2

configure readers — Update file-reader parameters. Omitted arguments keep their current values

Usage:

eubi configure readers [OPTIONS]

Optional arguments

--as_mosaic

Type: boolean flag

Default: default

Valid values: --as_mosaic to enable · --as_mosaic False to disable

Treat tiled acquisitions as a stitched mosaic (default False).

# Stitch tiled acquisitions into a single mosaic array on read
eubi configure readers --as_mosaic

--view_index

Type: int

Default: default

View index for multi-view formats (default 0). Pass 'all' or a comma-separated list (e.g. '0,2') to expose multiple views.

# Write each view separately
eubi configure readers --view_index all

# Select only the second view
eubi configure readers --view_index 1

--phase_index

Type: int

Default: default

Phase index (default 0).

# Select phase index 1
eubi configure readers --phase_index 1

--illumination_index

Type: int

Default: default

Illumination index (default 0). Pass 'all' or a comma-separated list to expose multiple illuminations.

# Write each illumination separately
eubi configure readers --illumination_index all

# Select only the second illumination
eubi configure readers --illumination_index 1

--scene_index

Type: int

Default: default

Scene / series index to read. Pass an integer, 'all', or a comma-separated list such as '0,2,4' (default 0).

# Always read scene 2 from multi-scene files
eubi configure readers --scene_index 2

# Convert each scene to a separate OME-Zarr group
eubi configure readers --scene_index all

# Convert only scenes 0, 2 and 4
eubi configure readers --scene_index 0,2,4

--rotation_index

Type: int

Default: default

Rotation index (default 0).

# Select rotation index 0
eubi configure readers --rotation_index 0

--mosaic_tile_index

Type: int

Default: default

Mosaic tile index. Pass an integer, 'all', or a comma-separated list (default None = all tiles).

# Write every tile separately (default)
eubi configure readers --mosaic_tile_index all

# Write only tiles 0 and 2
eubi configure readers --mosaic_tile_index 0,2

--sample_index

Type: int

Default: default

Sample index (default 0).

# Select sample index 0
eubi configure readers --sample_index 0

--force_bioformats

Type: boolean flag

Default: default

Valid values: --force_bioformats to enable · --force_bioformats False to disable

Force bfio tiled path even for natively-supported formats.

# Always route reads through the Java Bio-Formats reader
eubi configure readers --force_bioformats

--concat_views

Type: boolean flag

Default: default

Valid values: --concat_views to enable · --concat_views False to disable

Stack multiple views along the channel axis instead of writing separate OME-Zarr outputs (default False).

# Always stack views onto the channel axis
eubi configure readers --concat_views

--concat_illuminations

Type: boolean flag

Default: default

Valid values: --concat_illuminations to enable · --concat_illuminations False to disable

Stack multiple illuminations along the channel axis instead of writing separate OME-Zarr outputs (default False).

# Always stack illuminations onto the channel axis
eubi configure readers --concat_illuminations

configure concatenation — Update aggregative (concatenation) parameters. Omitted arguments keep their current values

Usage:

eubi configure concatenation [OPTIONS]

Optional arguments

--concatenation_axes

Type: str or int or None

Default: default

Axes along which to concatenate files. Pass a string of axis letters such as 'tc' or 'z', or an integer axis index. None disables aggregative mode entirely.

# Concatenate files along the channel axis
eubi configure concatenation --concatenation_axes c --channel_tag raw,mask

# Concatenate along both z and channel axes simultaneously
eubi configure concatenation --concatenation_axes zc --z_tag slices --channel_tag raw,mask

--time_tag