coherent WaveBurst Review Meeting Monday 15 Jan 2009 08:30 Pacific / 11:30 Eastern
Agenda: Thursday 15 Jan 2009 08:30 Pacific / 10:30 Eastern
- Selecting a telecon time.
- Status of cWB S5yr2-VSR1 analysis: low frequency.
- Status of cWB S5yr2-VSR1 analysis: high frequency [ wiki ].
- What needs to be reviewed.
- Division of tasks.
Dial-in number: 1 800 704 9896 International dial-in: +1 404 920 6472 Conference code: 5374 2349 #
Sutton (minutes), Yakushin, Klimenko, Gouaty, Salemi, Re, Prodi, Drago, Babak, Zanolin.
Everyone was urged to fill in the doodle poll for a regular telecon time.
On Thursdays there are conflicts with the IMR call (11:00 eastern, 30-60 min), the high-mass call (11:00 eastern, 60-90 min), and the omega call (11:00 eatsern, 60 min).
Status of low-frequency cWB S5yr2-VSR1 analysis
- [ web ] Sergei: Igor has posted a web page showing livetimes for each network combination, jobs, data, analysis pages, explanation of CVS versions, run parameters, post-processing cuts, etc.
- One HxL1 network has not been opened; all other opened.
- ULs to come from 4x, 3x, and H1-H2 combinations.
- Action item: Some entries on Igor's page need to be updated.
- Action item: All web pages need to have links to zero-lag events.
- Action item: Igor to post files for MDC runs in format requested for making combined upper limits.
- Action item: All input files (e.g., segment lists) need to be archived in a CVS.
Status of high-frequency cWB S5yr2-VSR1 analysis
- [ web ] Virginia: Walked through the Virgo wiki for the high-frequency cWB search.
- 1.28-6kHz search range.
- 5 networks: 4x, H1H2L1, H1H2V1, H1L1V1, H1H2.
- Opened boxes for multi-site networks in December -- no events. Only H1H2 remains to be opened.
- Half the lags used for tuning, the other half for background estimation.
- Analysis to date has only used Q9 SGs. Do we want more waveforms? Two more sets of MDCs have been generated: one with Q3,100 SGs and 2 SN waveforms, a second with WNBs and ringdowns.
- Did calibration tests using JW1 data.
- Web page contains suggested items for review.
Status of first-year review
Michele: Analysis review document circulated before the December meeting. Response positive -- everyone seems mostly pleased. The document is final except for:
- Segment checks -- to be finished by Sutton and Yakushin.
- Sensitivity studies -- only done for H1H2L1, but otherwise okay (Sutton).
- Characterisation and physical origin of low-frequency loud events in the background. Signals are low-Q, at 100Hz. Origin unknown.
What needs to be reviewed
Patrick collected a list of possible review items. These should be done for both the low- and high-frequency searches, were applicable.
- Review of new and modified codes. (Sergei: Heavier for HF than for LF.)
- Walk-through of all results pages.
- Data quality and segment checks. Check livetimes -- we've analysed all the data we intended to analyse, and the correct flags have been applied. (Sergei: Hardest review task!)
- Calibration tests: check the Trento study results and see if we need anything further.
- Loud event follow-ups (background, assuming no more zero-lag events turn up).
- Loud missed injection follow-ups.
- Analysis results checks: consistency of tuning and UL results (background distributions of rate vs. threshold).
- Auto-correlation plots.
- Double-check use of 2 weeks of data accidentally included in the first-year search (low frequency analysis only).
The goal is to have the review closed out for the March LVC meeting. (Patrick: I think this is optimistic.) Sergei: We need to present the paper to the review committee about 3 weeks earlier, to get their approval. This means we would have to finish the review by the end of February.
Division of tasks
Action item: Patrick will draft preliminary division of tasks for next week. All are invited to send suggestions for tests and to volunteer for tasks.
We will meet next week (during GWDAW) and those available will start the code review.