[INSTRM-1720] Add grouping to gen2 dither scripts Created: 25/Sep/22  Updated: 01/Mar/23  Resolved: 01/Mar/23

Status: Done
Project: Instrument control development
Component/s: PFS_kansoku
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: cloomis Assignee: Shintaro Koshida
Resolution: Done Votes: 0
Labels: EngRun
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Blocks
blocks INSTRM-1763 Prepare Gen2 abstract commands and op... Done
is blocked by INSTRM-1719 Pass iic groupIds to Gen2 Done
Story Points: 2

 Description   

We want to group all exposures in a dither sequence under a common id. From the Gen2 point of view, the mechanism will be as follows.

Before starting the dither loop, request a new group from iic getGroupId groupName=$NAME, where I recommend using "dither" for the $NAME.

The response will be an integer in PFS.GROUPIDS.$NAME.ID. That id should then be passed the exposure commands with groupId=$ID.

For example, in semi-Gen2 pseudo-code:

exec iic getGroupId groupName=dither

ditherGroupId = !PFS.GROUPIDS.DITHER.ID

for ditherPos in ditherPosList {
   exec icc scienceObject exptime=90 window=500,1000 groupId=ditherGroupId
}

If you need to restart a dither sequence (to add a new ring, for example), you can request the groupId for an existing group with iic getGroupId groupName=dither continue. If you do not add that continue you will get a new id.

[ I might add .VISIT or .DATE to the PFS.GROUPIDS.DITHER dictionary, which is why there is an additional .ID ]

This ticket will be blocked



 Comments   
Comment by Shintaro Koshida [ 25/Sep/22 ]

Thank you, Craig-san, for the detailed explanation. It looks very straightforward and possible to work on it as soon as the PFS status key PFS.GROUPIDS.DITHER.ID is implemented. Maybe I would ask Takagi-san if he can work on the scripting first, because the dithering code is created and developed mostly by him so far.

One question; do you have any idea in your mind about when this ticket should be closed? By the next run, or in this run if possible?

 

Comment by cloomis [ 17/Dec/22 ]

Eric convinced me that I was being too tricky: we will only need to request one groupId at a time, so there is no need to the nesting.

So I have switched to a flat namespace. For my testing, I see this on g2sim1:

PFS.GROUPID.DATE: 2022-12-16T22:40:21
PFS.GROUPID.ID: 20
PFS.GROUPID.NAME: cpl
Comment by yuki.moritani [ 04/Feb/23 ]

Just for recording; this is implemented as a part of INSTRM-1815.

Comment by yuki.moritani [ 01/Mar/23 ]

branch INSTRM-1815 was merged with this modification, so I close the ticket.

Generated at Tue Apr 08 20:48:09 JST 2025 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.