X-Pipeline Review Meeting Monday 13 Oct 2008 09:00 Pacific / 12:00 Eastern
Agenda: Monday 27 Oct 2008 09:00 Pacific / 12:00 Eastern
Review summit at Cardiff:
Nov 03 - 07 Nov 10 - 14 Nov 17 - 21 Dec 01 - 05
- Data quality flags and detectors: web page
- Code review: At line 527 of xtimefrequencymap.
Minutes:Attendance: Patrick, Gareth, Jolien, Xavi, Antony
- Ping on review summit.
- DQ flags - trying to settle on choices about whether to include
a detector in an analyis.
- Patrick's proposal: decide on network based on the amount of deadtime in cat2 and 3 vetoes -- if a few seconds, keep detector and suffer efficiency; if a few minutes, drop detector; only a couple of cases in between.
- Previously: if a DQ flag is on during a GRB onsource interval, that detector isn't used for that GRB.
- Jolien: doesn't like the implicit prior that GWB equally likely to be present during 180s window. DQ flag on at time of GWB is far more devistating than at beginning of window.
- Suggestion: apply cat 1 and 2 flags, not 3 ("detection search" flags). Cat 3 are not really derived for GRB searches -- designed for ULs for all-time searches. Don't use them.
- Different network for detection vs. upper limit. If cat 2 veto -- don't include detector in network for UL. Detection: do what is possible; report 50% efficiency as sensitivity; indicate what time was vetoed.