240220.md

@@ -16,7 +16,7 @@ Urja
 
 ## Focus Session
 
-[Virgo Final Call](https://docs.google.com/document/d/1DQkGEii_6lMEUSWIFzvNKLqWWw3_wF27GzayQInXAjc/edit?usp=sharing)
+On [offline injection filter and rank](https://docs.google.com/presentation/d/1PqxQf5te3dqPAIUzd0aTowFVj9pnPWSrrr4Sw0k8VZM/edit#slide=id.p)
 
 ## Action Items for Next Week
 
@@ -26,26 +26,67 @@ Urja
 
 * Announcements (5 minutes)
   - Please check the rota for next week's call
+    * Zach and Amanda
   - Confirmation of next week's focus session
 * Last week's East call
+  * Leo : Cort presented the patch to fix the segment bug to find the 1-IFO or 2-IFO times. Kip wasn't convinced that the patch fixed the bug for a strange edge case. 
+  * Cort : weird edge case where for 1 second it shows 2 IFOs but it actually just is for 1 IFO. The patch is from 2022. Chad will present it again and it will get merged tomorrow likely.
+  * Shomik : Make it as an action item
 * Quick updates (45 minutes)
   - Operations (5 minutes)
     - LL CBC operations
+      * Shomik : Will be in lock in about 2 weeks from now. Will be testing to see if we can process any data coming in from the detectors. Getting started on that this week. 
   - O4 Dev (30 minutes)
-    - Low latency integrated testing and Monitoring
+    - Low latency integrated testing and monitoring
+      * Cort : updated the way the test suite env is handled. Use make launch to launch the dag instead of condor_submit_dag. 
     - O4b online production lead team updates
       - Engineering run plans
+        * Keep running the commissioning break analysis as it won't collect background data. We will start burning in for new analysis with O4b. We will include virgo in filtering plans.
       - AllSky
+        * LL sent out a circular saying that all analysis need to participate from 22 Feb to 7 March uploading to gracedb-playground and running on MDC data. Will hopefully launch something tomorrow or on 22nd. 
       - EW
+        * Takes longer to run so we've already launched it and have LRs. Should start uploading in about 2 days. 
       - SSM
+        * Shio : Launched checkerboard on ICDS and CIT. Need to put in the last patch that Zach made and update the build for that. 
     - IMBH
+      * Shio : Testing the bank with offline runs. 
+      * Debnandini : I am working on [this ](https://chat.ligo.org/ligo/pl/fyzpxm958if1dpym19re3s46uy)task, the dag is at the injection filtering stage. We are also looking into mass model options/issues on the side.
     - GWTC-4
+      * Leo : Not much updates. Schimma is updated to have Pastro, FAR for each Superevent in a table format. Chad is working to adapt that change into the GWTC package. 
     - Computing
+      * Shomik : Nemo cluster are incorporating a similar plan as ICDS and CIT so we have dedicated nodes/slots. 
     - Likelihood ratio, background and foreground sampling
+      * No updates
     - Injection file format
+      * No updates
     - SNR optimizer
+      * Zach : Testing the changes today with all sky tag on gracedb test
     - Offline DAG and Ops
-  - Exploratory development/Misc projects (5 minutes)
+      * Cort : rank dags finished so the injection workflow has finished. There should be results soon. Working on fixing proxy issues but shouldn't be causing any issues.
+      * Ryan : Has the issue with abnormally high SNRs been brought to the R&P teams attention? 
+      * Becca : We should make sure it isn't a problem on our end before going to R&P
+      * Ryan : Cody plotted the actual injection strain and it looks totally fine. Not a problem with our pipeline. Wrong expected SNR value is stored. So that could be an us problem if R&P messed up the injection which means we'd have to do extra injection filtering. 
+  - Exploratory development/Misc projects (5 minutes). 
+    * Nothing here
   - Paper Updates
+    * Becca : Performance paper was accepted
   - Assigning sprint tasks to those who were not at sprint.
+    * Zach : checking on if there are places where anyone is blocked with their sprint tasks
 * AOB
+  * VIRGO
+    * 50 MPC sensitivity
+    * We should filter Virgo data 
+    * Blue dots - advok and we would've removed counts 
+    * Orange dots - we wouldn't have removed counts. Concern about this causing background contamination. 
+    * Similar number of events in O4a and O4b. 
+    * 3% of max will be in background. About 1 event contaminating it. We could change the threshold to about 1/1 month or 1/2 months. It would ensure that if there is a true GW event, it won't be causing an issue with background but does change our statistics. 
+    * Leo : Isn't convinced that that is the right thing to do. More comfortable if we had dtdphi veto deployed plus lowering this threshold and so this corner case won't happen as much. 
+    * Shomik : Does the dtdphi veto change the FAR or does it change whether the event will be uploaded or not? 
+    * Leo : It will prevent event from being uploaded. We won't remove the counts are it will likely be noise
+    * Shomik : So it will never end up on gracedb. 
+    * Leo : Without the veto, it will be uploaded but we will get ADVNOK and has to be retracted. This is the corner case that is concerning. More effort on our side to being up dtdphi veto and convince reviewers it's the right thing to do. 
+    * Shomik : We don't have review slots for this. Should make a decision for this by today or tomorrow. 
+    * Leo : Retrieve suspicious events and remove them once the count tracker 
+    * Shomik : ranking on V and use the count tracker as we did in O4a 
+    * Leo : If we have bad contamination, we can still go back and fix things
+    * Shomik : Update for tomorrow's LL call - we will be filtering Virgo data but won't mention about the count tracker. 
\ No newline at end of file