Open mm6654 opened 1 year ago
The populating issue is due to multi active sessions for the same annotator, brain and type. The new feature, state id, might cause the failure of deactivating sessions with empty state id.
I checked this page and got 486 points: https://activebrainatlas.ucsd.edu/admin/neuroglancer/annotationsession/?q=dk62 Whereas on this page I got double, or 972 points: https://activebrainatlas.ucsd.edu/activebrainatlas/admin/neuroglancer/markedcellworkflow/?q=dk62 The workflow page has an error on it. It is also counting inactive sessions, so the real count is 486 and that is what gets imported. When you select the import, it will show the inactive ones also. This is another error. I'll try to fix these later today.
This has been corrected. Please check and close. As for Kui's concern, to get a session, the code first looks for a session with the state ID, if none is found, it looks for one without a state ID. See code at line 252: https://github.com/ActiveBrainAtlas2/activebrainatlasadmin/blob/master/neuroglancer/annotation_manager.py
Does this mean the "save annotations" button will populate the Annotation Sessions and Marked Cell Workflow tables correctly?
Once tested and implemented on production server, yes. I’m not sure if fix is on webdev right now
Duane
From: Marissa Moreno @.> Sent: Thursday, December 22, 2022 10:57 To: ActiveBrainAtlas2/activebrainatlasadmin @.> Cc: Duane Rinehart @.>; Assign @.> Subject: Re: [ActiveBrainAtlas2/activebrainatlasadmin] Clicking "Save annotations" does not populate Marked Cells Workflow (Issue #154)
Does this mean the "save annotations" button will populate the Annotation Sessions and Marked Cell Workflow tables correctly?
— Reply to this email directly, view it on GitHub https://github.com/ActiveBrainAtlas2/activebrainatlasadmin/issues/154#issuecomment-1363233295 , or unsubscribe https://github.com/notifications/unsubscribe-auth/AGDPO3UYV67QF5WTJL4TDO3WOSP7NANCNFSM6AAAAAATGCSWLM . You are receiving this because you were assigned. https://github.com/notifications/beacon/AGDPO3Q6SJWQ7FR6BTIGFALWOSP7NA5CNFSM6AAAAAATGCSWLOWGG33NNVSW45C7OR4XAZNMJFZXG5LFINXW23LFNZ2KUY3PNVWWK3TUL5UWJTSRIFFA6.gif Message ID: @. @.> >
I checked the Active Brain Atlas portal and for DK62 Marissa Unaided Test and the actual number of annotations should be 258 not 486. If you made the fix is it through webdev? If so, how can I test it?
I did not update webdev, only https://activebrainatlas.ucsd.edu. I just updated webdev also, so that should read correctly.
I see, but the problem is that the number of marked cells being shown on both tables is incorrect. I only made 258 marked cell annotations in this Neuroglancer state, but the tables show that there are 486 marked cells.
When I look at this page: https://activebrainatlas.ucsd.edu/activebrainatlas/admin/neuroglancer/points/points-data/774 I get a count of 259. There must be one redundant entry. That page just parses the JSON data so there must be an error with the save annotations button. I'll look into it today.
I fixed the problem and put the updates on webdev. The numbers should match between the values in the annotation session page, the marked cell workflow page and the points page.
Ed to put on production server, Marissa to test on 12-JAN-2023
I deployed to production server, please test and close.
This issue was fixed on Active Brain Atlas, but not on the BrainSharer Portal
Marissa: Is this issue now correct on brainsharer.org (so we can close)?
For the latest QC on DK62, after I clicked save annotations, the "Marked cells workflow" table did not populate with the correct amount of annotations. Assuming the annotations saved correctly, how would I import the correct number of annotations into a new brain? Or will it import the wrong number of annotations?