X-Pipeline Review Meeting Monday 01 Dec 2008 09:00 Pacific / 12:00 Eastern
Agenda: Monday 01 Dec 2008 09:00 Pacific / 12:00 Eastern
- Announcements: H1H2 standard likelihood bug fixed; affected GRBs rerun.
- Jolien's "checklist" added to review summary page.
- Code review: Reviewing likelihood types in lines 706--1192 of xtimefrequencymap.
Minutes:Attendance: Patrick Sutton, Jolien Creighton, Gareth, Xavi
- H1H2 likelihood corrected and rerun.
- Before opening box todos: sort out injection issue, comprehensive calibration studies, go over xcondition again.
- Concern about null-stream cut. Residual signal if calibration is not correct. Cut is in comparison to not-null stream energy. Looking for significant amount of cancelation rather than complete cancelation. However, phase errors (e.g., time error ~1ms) would put one signal out of phase with other signal. V3 calibration: a few degrees (5 degrees) at a few kHz.
- Would like MDC study with miscalibrated injections. Don't need high accuracy estimates (large number of injections) per GRB but would be nice to have it for all GRBs because of the different antenna patterns. Can do this within the xpipeline injection engine.
- Try things out with one GRB.
- Population study: binomial test as Isabel did. Is it better to to use a single loudest event, loudest 10 events, ?
- Need a population plan.
- Question about review summit.