Introduction

The purpose of this page is to investigate the effect of using a composite detection statistic described in arxiv:1101.0584 to first identify interesting places to reconstruct SNR. The idea is that it can save floating point operations.

Method

This is preliminary and shouldn't be considered a final answer about whether or not this is a good idea.

Two gstlal_inspiral analyses were run. The first did not use a control peak time. The second did not, e.g.,

    channa@pcdev1:~/MDC_new/test$ diff -u Makefile ../test_2s/Makefile
    --- Makefile    2014-06-17 20:35:16.969869229 -0500
    +++ ../test_2s/Makefile 2014-06-17 20:32:55.203392803 -0500
    @@ -17,10 +17,10 @@
     IFOS = H1 L1 V1
     START = 966384015
     STOP = 967384015
    -TAG = T1200307_LV_gaussian_0s_w_zerolag_injections_40Hz_5ms_test
    +TAG = T1200307_LV_gaussian_2s_w_zerolag_injections_40Hz_5ms_test
     WEBDIR = ~/public_html/MDC_new/${START}-${STOP}-${TAG}
     NUMBANKS = 4
    -PEAK = 0
    +PEAK = 2
     AC_LENGTH = 351
     # additional options, e.g.,
     #ADDITIONAL_DAG_OPTIONS = "--blind-injections BNS-MDC1-WIDE.xml"

Results

The results are here:

The range plots are shown here:

Range in Mpc vs FAR for no control peak time

Range in Mpc vs FAR for control peak time of 2 seconds