Crash/Error using Biologically-inspired Dataset split #109
Unanswered
CarrieASims
asked this question in
Q&A
Replies: 1 comment 2 replies
-
|
Hi CarrieASims, This is a bug of some recent changes. I suggest that you export the dataset without using this option. If the problem remains, let us know. |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
TagLab is crashing when I try to run the Biologically-Inspired dataset split when Exporting a New Training Dataset
This is the error:
PS E:\5_Taglab\TagLab-2022.11.3> py taglab.py
TagLab 2023.5.17
Traceback (most recent call last):
File "E:\5_Taglab\TagLab-2022.11.3\taglab.py", line 4060, in exportNewDataset
new_dataset.setupAreas(mode.upper(), target_classes)
File "E:\5_Taglab\TagLab-2022.11.3\source\NewDataset.py", line 667, in setupAreas
val_area, test_area = self.findAreas(target_classes=target_classes)
File "E:\5_Taglab\TagLab-2022.11.3\source\NewDataset.py", line 397, in findAreas
landscape_number, landscape_coverage, landscape_PSCV = self.calculateMetrics([0, 0, map_w, map_h], target_classes)
File "E:\5_Taglab\TagLab-2022.11.3\source\NewDataset.py", line 291, in calculateMetrics
if self.frequencies[key] > 0.0:
KeyError: (0, 'Background')
PS E:\5_Taglab\TagLab-2022.11.3>
I have limited understanding but seems it has an issue with the class Background, which is set default as Empty in the TagLab interface.
Any assistance/enlightenment is most appreciated.
thanks
Beta Was this translation helpful? Give feedback.
All reactions