-
Notifications
You must be signed in to change notification settings - Fork 8
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Command errored: Open Multiple Images and Labels... [ilastik .h5] #167
Comments
Hi @ovalerio, sorry that ran into issues with the ilastik output in MoBiE - it's so nice when it works... Do you think you could share the images with us? I hope we can resolve this quickly Cheers |
Hello Dominik: Certainly. =) In Github I am restricted to max 25 MB file upload size. For context, I am creating the chunk files using h5py:
Thanks! |
Hello @ovalerio, thank you so much for providing the data, and apologies for the delayed response! Also, your details on how you created the h5 files in the first place was super helpful! You're very close, even more of a pity it took me so long to reply :/ The reason MoBiE cannot load the data is that it is missing some attribute that helps MoBiE understand the meaning of the axes in your data. Your data has 3 dimensions, so I assume it should be likely either "tyx" or "zyx". From the exported files I can see that you treated your data as 3D (3 spatial dimensions) ilastik. Generating the compatible metadata would entails somehting like this: import h5py
# optional, in case you don't mind adding this dependency
import vigra
# then open your h5 file
with h5py.File("chunk0.h5", "a") as f:
f["exported_data"].attrs["axistags"] = vigra.defaultAxistags("zyx").toJSON()
# if you don't want to use vigra, so also no need to import at the beginning,
# no problem, you would alternatively do:
axistags = """
{
"axes": [
{
"key": "z",
"typeFlags": 2,
"resolution": 0,
"description": ""
},
{
"key": "y",
"typeFlags": 2,
"resolution": 0,
"description": ""
},
{
"key": "x",
"typeFlags": 2,
"resolution": 0,
"description": ""
}
]
}
"""
with h5py.File("chunk0.h5", "a") as f:
f["exported_data"].attrs["axistags"] = axistags
# for the rest of the input files, you'd do the same. The exported probabilities have this metadata attached. also, reference to our jupyter notebook for h5 conversion: https://github.com/ilastik/ilastik/blob/main/notebooks/h5convert/convert_to_h5.ipynb Hope this helps. And please if you have any further questions, let me know. Cheers |
Hi @ovalerio, Could you please share the files and how you exactly open them in MoBIE such that we can try reproducing this error? Thanks! |
Dear
MoBIE
developers,Thank you for creating this Plugin.
I learnt about it while researching about QC tools for Pixel Classification jobs in
ilastik
.As you can see in the
FIJI
log, I tried to open my images withMoBIE
:The problem is that I got an IllegalArgumentException:
I am using
ilastik
in headless mode. Here is the log for one of the substacks:Thanks in advance for your help.
The text was updated successfully, but these errors were encountered: