Visualization of S5R4 results
19 Oct 2008 18:33:27 UTC
Topic 193987
(moderation:
Hi!
In order to get something like an end-to-end test of the LIGO hardware and software, some fake signals (including also fake pulsar signals) are deliberately inserted into the LIGO interferometers. Checking the results that my computers have returned to E@H, I now noticed that I was lucky enough to get a workunit that is sufficiently "close" to an injected pulsar so that the E@H app shows a noticeable response to it.
Look at this:
This is a visualization of the result h1_0575.00_S5R4__244_S5R4a_0_0 , showing column nr 5 as a measure of "significance" for a candidate.
It's better visible if you visualize column nr 8 of the result file instead (the column, along with others, was added in S5R4, which explains why S5R4 resultfiles are bigger in size than S5R3 results):
(now, this stands out pretty well).
Looking at all my S5R4 results so far and filtering for those which have high values in column 8, (a poor man's post processing :-) ) I get this picture:
The green peak is caused by workunits near the injected pulsar described above. Most of the blue speckles are at exactly 256.00 Hz, I guess this is an artefact from the sampling frequency of the "gravitational wave channel" which is 2^14 Hz AFAIK.
I'm not sure where the red peak is coming from, maybe from a "calibration line" near 1144 Hz (as mentioned in the S4 E@H result paper).
CU
Bikeman
Visualization of S5R4 results
)
That is so cool! It's like there is a finger pointing at it. Gives a good visual 'needle in a haystack' feel. Well done. Thanks for that. :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Really nice, only my machine
)
Really nice, only my machine don't displays graphics at E@H. Windows Vista 32 bit, AMD Turion64 x 2, BOINC 6.2.19. Laptop, not overclocked. I remembered tha in some older Power Apps. there was something about no graphic awailable, so I thought it is an unsolved problem. Looks like not. Any suggestions?
Christoph
Greetings, Christoph
RE: Really nice, only my
)
The diagrams produced here are not generated by the E@H app but by an open source software called "Topcat". The procedure to intercept and visualize E@H results is a bit complicated and described here.
CU
Bikeman
RE: RE: Really nice, only
)
Hello!
In S5R3 result file was five columns, in S5R4 file - nine. If I "reduce" file to five columns, I get a right data? :)
Thank you!
RE: RE: RE: Really
)
Yes,the first five columns in S5R4 results still have the same meaning as in S5R3.
CU
Bikeman
I learned how (maybe
)
I learned how (maybe correctly) to use TopCatLite v3.1 to create single gifs showing multiple WU's. I think I am showing how the different WU's are searching different areas of the sky, however it doesn't show if all of the WU's are using data from the same time or frequency range.
Here are two gif versions of the same 5 WU's
NOTE: This web space, and the gif's, may not be available for more than a couple of months.
Einstein_WU_Combo.gif
Einstein_ALT_WU_Combo.gif
RE: I learned how (maybe
)
Beautiful, isn't it !
As for
the different WU all cover the same time interval: the second part of the S5 science run. That's why the datafiles are comparatively large, they have to cover many months of data (but only for a small frequency bandwidth).
If the WU names are equal except for the sequence number (following the double underscore __ in the WU name, and if the sequence numbers are reasonably close together, the data will be from the same frequency range, but to make sure, you can use Tomcat to color the data points according to the first column (the frequency) to check.
CU
Bikeman
Hi I have some pictures
)
Hi
I have some pictures created with Topcat.
It is S5R4_0589.65 sequences
There is no significant peak in column 5 view
but ...
column 9 side 1 view
The circle in the graph may be due at Milky-way
column 9 side 2 view
column 8 view
You can see a peak in this graph
column 8 of h1_S5R4_0589.65_231 WU
Here you are the data file
http://dl.free.fr/us8j9owqt
good bye
Romain
I'm having trouble
)
I'm having trouble downloading from the link you posted, and as I write this, the images embedded in your post sometimes fail to load :-(. But I've noticed as well that there seems to be a little "bulge" when you visualize columns 8 and 9 in almost all results near the equator. Geometrically this seems to coincide with the ecliptic plane and not with the galactic plane, tho. I guess this is an artefact of the way the search algorithm works for different places of the sky.
Here's an example:
Or......we've just tuned in on a flying saucer commercial on some alien Gravitational Wave band :-).
CU
Bikeman
RE: Or......we've just
)
Well, they could be bedding down on Pluto as we speak, having scooted in with the warp drive. They're probably wondering if we'd spotted them. :-)
It does look like a ringed planet, or an accretion disk, or some ugly cosmic merger, or an imminent nuclear instability, or the old fashioned type of child's spinning top ...... :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal