Avoid writing into the labnotebook before the sweep can be aborted
Opened this issue · 0 comments
t-b commented
File: issue-1743\Pvalb-IRES-Cre;Ai14-646904.13.03.02_original.pxp
MIES version:
Release_2.6_20220606-39-g0882798c
Date and time of last commit: 2022-09-02T17:27:13+02:00
Submodule status:
f3493461480eaafa97939940693ed693860048a6 Packages/IPNWB (f349346)
56f1da31d786a593892280c2d2b2b1190def0fb5 Packages/doc/doxygen-filter-ipf (heads/main)
a8d4707e7c72555baefbb9a4d52de920e1d8ebb9 Packages/unit-testing (a8d4707)
Dashboard with main 75613dd (Merge pull request #1672 from AllenInstitute/bugfix/1672-warn-about-empty-wb-epochs, 2023-05-26) reveals that the "DAQ stop reason" for sweep 41 can not be queried.
Note the 42
entry in line 1649
with the later 41
entries from TP again.
Breakdown of what happened (guess):
- TP was running after sweep 41 was stopped manually (
DAQ stop reason
from the logbook viewer) - The entry in line
1649
isPSQ_RAMP version
that is called fromPRE_DAQ_EVENT
/PRE_SET_EVENT
inPSQ_Ramp
- We do have in the history
(ITC18USB_Dev_0) Please select a different indexing end setimset for DA channel 0 referenced by headstage 0
- We then have TP entries for sweep 41 again
So this looks like PSQ_Ramp was started but then aborted due to the history message. The history does not have a timestamp, so this is only a guess. And as TP ran again after that try we have a messed up labnotebook.
To fix that:
- Don't write into the labnotebook from analysis function events which are before the point where we definitly start DAQ. These events are
PRE_DAQ_EVENT
/PRE_SET_EVENT
/PRE_SWEEP_CONFIG_EVENT
. For these events we also pass in the future sweep number of the next sweep.
Should be coordinated with #1235.