Data Simulation Software
DMTGen
DMTGen is a stand-alone
off-the-shelf driver program for the DMT data simulation package. The
user can specify
one or more pre-recorded or generated signals to be combined and
written to frames as raw data channels. DMTGen can write simulated
event parameters, detector descriptions and history information to the
output frames in addition to the output channels. More information on
the features and use of DMTGen can be found here.
MatchTrig
MatchTrig is a program that compares trigger results of DMT monitors to
the generated event parameters recorded in the Simulated event frames.
MatchTrig produces a set of plots summarizing the trigger efficiency
and esolution including the following:
- Global time difference
- Efficiency vs. amplitude
- Time offset vs. amplitude
- Time resolution versus amplitude
And for each selected parameter, plots are made of
- Efficiency
- Time offset
- Time resolution
- Fractional amplitude offset, and
- Amplitude resolution
versus the parameter value.
MatchTrig can be run on any trigger set for which the data frames
with FrSimEvent structures are available. There is no need to make any
changes to a monitor.
Running MatchTrig
The MatchTrig command line is:
MatchTrig {-infile <files>
| -inlist <list-file>} -ttable <trigs> [-debug
<debug-level>] \
[-maxdiff <max-dt>]
[-tlimit <histo-time-limit>] [-amin <ampl-min>] [-amax
<ampl-max>] \
[-histo
<hist-file>] [-par <par-name>,<min>,<max>]
[-select <trigger-id>/<sub-id>]
The command line arguments have the following meaning:
-infile
|
Read simulated data from frame
file <files>
|
-inlist
|
Read simulated data fom frame
files listed in <list-file>
|
-ttable
|
Read triggers from xml file <trigs>
|
-debug
|
Print optional debugging
information
|
-maxdiff
|
Match trigger to events if the
time separation is less than <max-dt>
seconds.
|
-tlimit
|
Histogram time offsets less than
<histo-time-limit> seconds.
|
-amin
|
Histogram lower limit for
amplitude histos.
|
-amax
|
Histogram upper limit for
amplitude histos. |
-histo
|
Store raw hitograms in ROOT file
named <hist-file>
|
-par
|
Print plots of efficiency,
average dt and amplitude offsets, and dt and Amplitude resolution
versus generated parameter <par>
value in the range <min> to <max>.
|
-select
|
Select triggers with ID <trigger-id>
and Sub-ID <sub-id>.
|
As of this writing, MatchTrig is not available in the DMT release, but
may be found in my Caltech home directories (e.g. on
saiph.ligo.caltech.edu and dmt[0-3].ligo.caltech.edu) at:
/home/jzweizig/work/tools/<platform>/bin/MatchTrig
Example - Triggers from PSLmon
As an example of the use of MatchTrig, I injected a bunch of
sine-gaussians into a continuous gaussian white noise background. In
this sample, the The DMTGen configuration file was as follows:
Parameter StartGPS 730000000
Parameter EndGPS 730100000
Parameter FilePrefix /usr1/store/S3MC/L1-SGGW
Source SG
SinGauss(A=flat(0,100),F=power(-1.0,100,2000),Q=flat(1,10),Phi=flat(-90,90))
\
-rate 0.2
-simdata -simevent
Source GS WhiteNoise(A=2.0)
Channel L1:LSC-AS_Q SG GS
I then generated triggers with PSLmon
on the
simulated L1:LSC-AS_Q
channel using the
following configuration:
Parameter TrigEnable 1
Parameter Stride 1.0
Channel L1AsQ L1:LSC-AS_Q
Glitch gTool1 L1AsQ -sigma 6.0 -trigger
I then ran the resulting trigger and data through MatchTrig with:
MatchTrig -ttable $GTable
-infile "$DATA_DIR/*.gwf" -histo MatchTrig1.file \
-par
Phi,-90,90 -par F,0,2000 -par Q,0,10 -select Glitch/gTool1 \
-maxdiff 0.01
-tlimit 0.1
The resulting plots are shown here. Note
that all three steps were run for 100,000 seconds of simulated data in
~4300s
seconds using a single processor on saiph (750MHz UltraSparc III?), i.e. a factor of
>20 faster than real time!
Last Update: January 30, 2004
Please send comments or suggestions to: John Zweizig