Note
|
These files are conlog output of all changes between the start of E7 and the end of E7.
We no longer use the conlog output of clean detect segments.
These files contain segments during which the instruments were in a clean detection mode. Segments are of duration ≥60 seconds.
Clean detection is defined such that:
The format is: start_time duration_in_seconds duration_in_hh:mm:ss
These files contain lock segment information, and were generated from SegGener meta-db segments:
These files contain raw lock segments, i.e. the identical data as in the previous section, but culled to only include segments of duration 60 seconds and above, and formatted differently.
The format is: start_time duration_in_seconds duration_in_hh:mm:ss
These files contain lock segments that are "clean", i.e. where the interferometer was in lock, and no controls were being changed. There is a recovery time of 60 seconds, i.e. the interferometer is not considered to be in a clean state until 60 seconds after the end of the previous segment. No recovery time was applied for changes in Epics controls. The segments are ≥60 seconds long.
The 60 second recovery time, and 60 second minimum segment duration, are arbitrary: no attempt has been made to see how varying these two numbers affects the statistics of the clean lock segments. There are 1087 clean H1 locks, 732 for H2, and 576 for L1.
The format is: start_time duration_in_seconds duration_in_hh:mm:ss
These files contain a sampling of clean lock segments that were computed by manually going through the raw segment data from GUILD and the conlog output. The same rules regarding recovery time and minimum segment length as mentioned in the previous segment were applied.
The output of coincidence segments was also validated and verified.
These files contain coincident clean lock segments, in all combinations of interferometers. The segments are ≥60 seconds long.
The format is: start_time duration_in_seconds duration_in_hh:mm:ss
H1:
No. of segments = 1087
Total lock time = 961002 s = 266:56:42
Max = 14687 s = 4:04:47
Mean = 884.086476541 s
Std. dev. = 1406.56939798 s
H2:
No. of segments = 732
Total lock time = 616204 s = 171:10:04
Max = 26687 s = 7:24:47
Mean = 841.808743169 s
Std. dev. = 2008.59576047 s
L1:
No. of segments = 576
Total lock time = 890724 s = 247:25:24
Max = 14294 s = 3:58:14
Mean = 1546.39583333 s
Std. dev. = 1876.52889925 s
H1 & H2:
No. of segments = 917
Total lock time = 501577 s = 139:19:37
Max = 13693 s = 3:48:13
Mean = 546.976008724 s
Std. dev. = 978.646982559 s
H1 & L1:
No. of segments = 1015
Total lock time = 677035 s = 188:03:55
Max = 10731 s = 2:58:51
Mean = 667.02955665 s
Std. dev. = 828.245908597 s
H2 & L1:
No. of segments = 692
Total lock time = 402689 s = 111:51:29
Max = 6611 s = 1:50:11
Mean = 581.920520231 s
Std. dev. = 819.075848415 s
H1 & H2 & L1:
No. of segments = 752
Total lock time = 333533 s = 92:38:53
Max = 4762 s = 1:19:22
Mean = 443.527925532 s
Std. dev. = 567.033232104 s
The clean lock segments and coincident segments were generated using this set of Tcl scripts (which should run if you have LigoTools installed): segments2.tgz (gzipped tar ball).
I also have a script that will take Guild's text output of segments (in the form start_time end_time) and convert them to a file of contiguous segments. (The segments from the LDAS database are at most 200 seconds long, so long segments get broken up.) The script is guild2segments.tcl, or you can view it here.
The (debug) output from the script is here: segments2.out.