Example Live Scripts 000011 & 000207 run sporadically in the MATLAB on DandiHub environment (R2024b)

[vijayiyer05]

As of the last working image provisioning MATLAB on DandiHub (as shown HERE; before the Nebari upgrade removed MATLAB images for now), there was an unresolved issue with running the example live scripts consistently on DandiHub.

I reported to @aranega that running the 2 example live scripts presently there (for Dandisets 000011 & 000207) were sporadic to run in my testing. He did further assessment and determined(*) the looped 'scatter' operation in the 000011 example was causing figure operations in MATLAB to fail thereafter (for either example). This could be remedied in his hands with a smaller loop (2 iterations vs 20).

Addressing/advancing this issue will require assessing the behavior against the scientific intent of the 000011 example, i.e., reducing the iterations should not be done as a "fix" unless it's scientifically sound. It is also hoped that any causing/contributing factors that appear to be on the side of MATLAB or the MATLAB instance on DandiHub specifically are brought forward.

This Issue is created with the intent to address in tandem with DandiHub Issue 276, Note this Issue, as described above, was reported against R2024b. It's possible this Issue will present differently (or not at all) on R2025a (what DandiHub Issue 276 aims to serve successfully).

(*) corrections/additions to my summary from @aranega are welcomed

[aranega]

@vijayiyer05 Thanks Vijay for the summary!

If I remember well, there was also the fact of clearing the variables from the environment between the execution of each script that fixed some issue; i.e: I think it allowed 000011 to be run with the right number of iterations after having executed 000207. This should be confirmed though. I, unfortunately, don't have a valid license at the moment, so I can't confirm if the fact of clearing the variable from the env was enough to remove the problem on 000011 (after the execution of 000207) or if this was only fixing the issue we encountered with MatNWB and the underlying schema version (problem that was seen only after the execution of the 2 scripts)

[vijayiyer05]

Thank you @aranega for this extra detail from your memory banks. It's good to have all we know/recall documented here on the Issue.

[SDV96]

IS THIS ISSUE SOLVED I CAN IMPROVE IT By..... if @vijayiyer05 approves assign within 1 WEEKS {WORKING NOW}