Closed KeremKurban closed 1 year ago
Hello, just use ' = []' for the regex thing, so it will not check anything. I tried locally, and S* creates your bug for some weird regex reason. This regex I added to be able to check that mtype names are of specific format, but it is not very useful, I should make it optional
Describe the bug
I was running the workflow for non-cortical region i.e. CA1 with layers SP, SO, SR, SLM so i set the mtype_regex as the following:
I assumed * would take care of all the rest but perhaps i had to be more specific if the code expects it such as this one
mtype_regex = ["L[1-6]_[A-Z]+$", "L[1-6]_[A-Z]+:[A-C]$", "L23_[A-Z]+$", "L23_[A-Z]+:[A-C]$", "L[1-6]_[A-Z]+-[A-Z]+$"]
I also added manually layer and mtype columns to dataset.csv after initialization. (Note in the documentation layer was not mentioned but it seems to be a mandatory field since i got an error asking for that column)
After pipeline was run, what i see in neurondb.dat is something like
where you can see mtype is S which is similar to S* field i put in the luigi.cfg
Is there a specific way of formatting we need to put the regex ? I see for every use case like L2-X , L23-Y L5:TTPC:A you put a different regex. in our case the mtypes are
SP_PC, SP_PVBC, SOOLM .. (i.e. {layer}{celltype} ) but it doesnt mean in the future for other regions/mtypes it will be like this as the nomenclature in literature will be set in such way.
Optional link from https://morphology-workflows.readthedocs.io which documents the behavior that is expected
No response
To Reproduce
Error
Additional context
sample dataset.csv first column looks like
morphology_workflows Version in Use
0.7.0
Python Version
3.10.8
Operating system
Linux