got error setMinMax

[ramdhan1989]

Hi I got this error :
File "C:\SuperHuman-main\SuperHuman-main\scripts\utils\lmc.py", line 19, in mc_baseline
costs = feats.compute_affinity_features(rag, affs, offsets)[:, 0]
File "c:elf\elf\segmentation\features.py", line 84, in compute_affinity_features
numberOfThreads=n_threads)
RuntimeError:
Precondition violation!
UserRangeHistogram::update(): setMinMax(...) has not been called.
(D:\bld\nifty_1641634469613_h_env\Library\include\vigra/accumulator.hxx:5935)

any idea to solve this?

thanks

[constantinpape]

Can you please share more information about your set-up:

• What operating system are you using?
• Which version of vigra do you use? (vigra is installed in the conda environment for elf)

[ramdhan1989]

Hi,
I used windows 11 and this is my vigra version 1.11.1.

Please advise
thank you

[constantinpape]

Unfortunately there seem to be some issues with this function on windows: #90
This is a bit more tricky to fix on my end and I currently don't have time for this.

You could try to replace it with boundary features. Can you describe what problem you want to solve and what shape the affinities have?
(Alternative: it works on Linux)

[ramdhan1989]

Hi,
Thanks for your information.
I want to use solution from this https://github.com/weih527/SuperHuman for my own case to segment seismic data as follow :

Thank you

[constantinpape]

Ok, you would need to update a few things here:
https://github.com/weih527/SuperHuman/blob/main/scripts/utils/lmc.py

• use compute_boundary_features on the first affinity channel to compute the features for edges between planes.
• use compute_boundary_features on the max of the second and third affinity channels for edges in-plane.

Here's a PR to implement that idea: weih527/SuperHuman#2

I have not tested this, so you would need to make sure that it works and potentially fix or report errors.