v0.4 axes transforms

[will-moore]

WIP:

Test napari https://minio-dev.openmicroscopy.org/idr/v0.4/2022-01-05/idr0062/6001240.zarr.
To show scale-bar, open the napari terminal and (on latest napari):

viewer.scale_bar.visible = True

(This was previously viewer.scale_bar_visible napari/napari#1720)

This PR requires https://pypi.org/project/ome-zarr/0.3a1/

To do:

• handle axes in v0.4 format
• handle transform scale to set pixel sizes
• affine transform ?!?

Not sure yet how to convert affine transform in the spec into what napari is expecting? https://napari.org/api/stable/napari.Viewer.html#napari.Viewer.add_image

- scale (tuple of float or list) – Scale factors for the layer. If a list then must be a list of tuples of float with the same length as the axis that is being expanded as channels.

- affine (n-D array or napari.utils.transforms.Affine) – (N+1, N+1) affine transformation matrix in homogeneous coordinates. The first (N, N) entries correspond to a linear transform and the final column is a length N translation vector and a 1 or a napari Affine transform object. Applied as an extra transform on top of the provided scale, rotate, and shear values.

[will-moore]

Without scale metadata, scale-bar shows pixels.
NB: size Z is 235.

Using pixel sizes in microns to add:
"scale": (0.5002025531914894, 0.3603981534640209, 0.3603981534640209)
to the metadata fixes the scale-bar to show correct size (compare with IDR image below), but now there are 470 Z-sections and if you go above 235 on the Z-slider you just get black.

See IDR https://idr.openmicroscopy.org/webclient/img_detail/6001240/

[will-moore]

You can also set units with viewer.scale_bar.unit='um' for microns.

[will-moore]

If I set the scale for only the X and Y axes, then I get a correct scale-bar without breaking the Z-slider:

"scale": (1, 0.36, 0.36),

but this distorts the 3D view:

[imagesc-bot]

This pull request has been mentioned on Image.sc Forum. There might be relevant details there:

https://forum.image.sc/t/napari-non-integer-step-size/31847/8

[will-moore]

Also see https://forum.image.sc/t/setting-scale-bar-units-in-other-than-pixels-real-coordinates/49158/12

[jni]

@will-moore what version of napari are you on? Do you think you would be able to make a reproducible example with just NumPy arrays? This works fine on main:

import napari
import numpy as np

_ = napari.view_image(
    np.random.random((10, 256, 256)),
    scale=(0.5, 0.3, 0.3),
)
napari.run()

I get a slider with 10 positions from 0 to 9 inclusive. The broadcasting of the scale doesn't seem to be the issue either, as the sliders also seem fine with this:

import napari
import numpy as np

_ = napari.view_image(
    np.random.random((3, 10, 256, 256)),
    scale=(0.5, 0.3, 0.3),
)
napari.run()

[will-moore]

@jni Thanks for that. I'm using napari 0.4.12 and your sample helped me to track down the issue. I was only applying the scaling to the image layers and not the labels layer so the sizeZ of the scaled image in real world coordinates is only a portion of the sizeZ of the labels layer.

[will-moore]

Looks like it's not possible to handle units yet - see napari/napari#1701

[will-moore]

Currently the tests are failing because the code now depends on ome/ome-zarr-py#124
cc @sbesson Do we need to do a pre-release of ome-zarr-py first and then update to that in the requirements here?

[sbesson]

cc @sbesson Do we need to do a pre-release of ome-zarr-py first and then update to that in the requirements here?

Yes that's one of the driver for the pre-release. The other one being ome/omero-cli-zarr#93 which is only passing because there are no tests.

I propose we review how close we are later today on signing off on the two open HCS ome-zarr-py and then decide on a 0.3.0 pre-release?

[jburel]

tested using ome-zarr==0.3a1.
Scalebar is at 25microns as expected.

[joshmoore]

Few minor points but happy to get this in for the meeting. I assume we keep it at a pre-release version until ome-zarr-py is ready.

[joshmoore]

Just a reminder that here we should probably think of a proper model (even if just using @propertys) as the mode

[sbesson]

@will-moore can any of the three points raised in #31 (review) be addressed/resolved quickly? Otherwise, I agree to get this merged and tagged as a pre-release in preparation of the community call.

[codecov-commenter]

Codecov Report

Merging #31 (b3abf92) into main (fcff2c8) will decrease coverage by 6.84%.
The diff coverage is 36.66%.

@@            Coverage Diff             @@
##             main      #31      +/-   ##
==========================================
- Coverage   92.77%   85.92%   -6.85%     
==========================================
  Files           3        3              
  Lines         180      199      +19     
==========================================
+ Hits          167      171       +4     
- Misses         13       28      +15     

Impacted Files	Coverage Δ
napari_ome_zarr/_reader.py	76.85% <36.66%> (-11.92%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fcff2c8...b3abf92. Read the comment docs.

[sbesson]

Thanks @will-moore. Merging and creating a pre-release