Differences between revisions 59 and 60
Revision 59 as of 2019-07-16 13:58:49
Size: 3440
Comment:
Revision 60 as of 2019-07-16 14:13:32
Size: 3680
Comment:
Deletions are marked like this. Additions are marked like this.
Line 91: Line 91:
== Online analysis ==

 * Online matched filtering
  * Using low latency data files dumped from shared memory
  * About 1 minutes deley
  * Template: (30,30)Msun, no spin, Waveform: IMRPhenomD
   * using pre-computed PSD
 * Plots

Engineering Run 2 (JST 2019/7/13)

  • JST: Jul. 13, 2019, 12:00 12:37 - 18:00 JST

  • GPS: 1247022018 1247024238 - 1247043618

  • KAGRA full data
    • m31:/data/full : Full KAGRA data
  • Low latency data
    • m31:/data/ll_merged/ : Low latency data packed in 4096 seconds
    • hydra:/data/ll_merged/ : Low latency data packed in 4096 seconds
    • hydra: /data/ll-sharing/shm/llhoft/ : Low latency data (1 second) dumped from shared memory (retained only for 8192 seconds)
  • channel name
    • From ER2 (DetChar)

    • K1:CAL-CS_PROC_MICH_DISPLACEMENT_DQ (We used this channel.)
    • K1:CAL-CS_PROC_IMC_FREQUENCY_DQ
  • Low latency data channels
    • K1:DCS-CALIB_STRAIN_C10 16384
    • K1:DCS-CALIB_STATE_VECTOR_C10 16
    • K1:CAL-CS_PROC_IMC_FREQUENCY_DQ 16384
  • lock flag
    • K1:GRD-LSC_LOCK_STATE_N = 157
  • data quality vector
    • definition

    • We cbc group should check the second bit is 1.
  • CBC lock segments
    • We checked lock flag of each data and collected the locked data. We considered lock segments by file units for convenience. If compared with lock segment(DetChar), this lock segment is small slightly.

      #

      start GPS

      stop GPS

      duration

      1

      1247024288

      1247043616

      19328

Off line analysis

Check point

  • Check the available locked data (done)

    • Summarize the data length (done)

  • Estimate power spectrum density (done)

    • Summarize information in pdf (done)

  • Generate lock list and segment list (done)

  • Generate template bank (skipped)

    • How many waveforms we need? (skipped)

  • Search signals by matched filter and chi^2 (done)

  • Summary the output of search (done)

    • histogram of SNR (done)

    • glitch list(GPS, JST, mass, ISCO frequency, duration) (not yet)

  • Summary of locked data CBC locked segment

  • Averaged PSD JGWdoc

    • Power spectrum density (PSD) is averaged over 1 sec chunks by taking median.
  • Calculation time to search for signals
    • Read 500 sec data : < 1 min

    • Estimate PSD : 2~3 min.
    • matched filter+chi^2 with single waveform (not bank) : < 1 min

  • Histogram of detection statistics JGWdoc

    • If data is Gaussian, <snr>=1.

  • Time series of detection statistics JGWdoc

    • waveform : 1.4 - 1.4 Msun
    • Beginning and ending 50 sec data were removed from the result.

Online analysis

  • Online matched filtering
    • Using low latency data files dumped from shared memory
    • About 1 minutes deley
    • Template: (30,30)Msun, no spin, Waveform: IMRPhenomD
      • using pre-computed PSD
  • Plots

KAGRA/DAWG/CBC/ER2 (last edited 2019-08-02 14:46:56 by HirotakaYuzurihara)