How do we know this software works?

Message boards : SETI@home Science : How do we know this software works?
Message board moderation

To post messages, you must log in.

AuthorMessage
S\

Send message
Joined: 18 Oct 11
Posts: 28
Credit: 4,719,993
RAC: 0
United States
Message 1807879 - Posted: 8 Aug 2016, 6:00:51 UTC

Have they fed false positives into the system to see if they come out positive?
ID: 1807879 · Report as offensive
Profile Wiggo "Socialist"
Avatar

Send message
Joined: 24 Jan 00
Posts: 12615
Credit: 169,679,224
RAC: 87,266
Australia
Message 1807880 - Posted: 8 Aug 2016, 6:09:08 UTC

Could you please elaborate?

Cheers.
ID: 1807880 · Report as offensive
Profile William Rothamel
Avatar

Send message
Joined: 25 Oct 06
Posts: 3335
Credit: 1,322,163
RAC: 0
United States
Message 1807927 - Posted: 8 Aug 2016, 11:43:01 UTC - in response to Message 1807879.  

Let's assume that the software works as intended. A better question is : What is it supposed to do and what is being done with the results of all the processing? Also: what methods and by whom are applied to the results of this processing??
ID: 1807927 · Report as offensive
Profile Bob DeWoody
Avatar

Send message
Joined: 9 May 10
Posts: 2848
Credit: 1,215,663
RAC: 134
United States
Message 1807935 - Posted: 8 Aug 2016, 12:32:42 UTC

I have always assumed the brainiacs at Berkeley knew what they were doing when creating the software.
Bob DeWoody

My motto: Never do today what you can put off until tomorrow as it may not be required. This no longer applies in light of current events.
ID: 1807935 · Report as offensive
rob smithProject Donor
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 15235
Credit: 253,188,634
RAC: 324,659
United Kingdom
Message 1807939 - Posted: 8 Aug 2016, 13:25:33 UTC

The software we run uses proven techniques for extracting signals from noise.
As others have said there are stages beyond our initial screening - these are just as vital, but are the ones that are currently proving to be "rather problematic" in both their implementation and execution.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1807939 · Report as offensive
S\

Send message
Joined: 18 Oct 11
Posts: 28
Credit: 4,719,993
RAC: 0
United States
Message 1807943 - Posted: 8 Aug 2016, 14:39:36 UTC - in response to Message 1807939.  

The software we run uses proven techniques for extracting signals from noise.
As others have said there are stages beyond our initial screening - these are just as vital, but are the ones that are currently proving to be "rather problematic" in both their implementation and execution.


I understand the methods are sound. But I'm curious if the software has been tested against a positive result.

I was part of finding Mersenne primes, and the last one was found, not because the automated software correctly identified it and alerted the community, but because the individual tester happened to be glancing at his results manually.

I don't mean to spread doubt, but I feel this is a valid question. Do we know S@H works?
ID: 1807943 · Report as offensive
rob smithProject Donor
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 15235
Credit: 253,188,634
RAC: 324,659
United Kingdom
Message 1807945 - Posted: 8 Aug 2016, 14:55:09 UTC
Last modified: 8 Aug 2016, 14:55:33 UTC

Unlike tasks like the Mersenne primes search we are not looking for things that are recognisable, but for things that don't match the pattern of human and "natural", or should that be "stellar"?, events. For example, our screening gets rid of known terrestrial RF sources. One of the test cases for any new application is to make sure that these are ignored, and only "non-terrestrial" signals are kept for further examination.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1807945 · Report as offensive
S\

Send message
Joined: 18 Oct 11
Posts: 28
Credit: 4,719,993
RAC: 0
United States
Message 1807949 - Posted: 8 Aug 2016, 15:19:52 UTC - in response to Message 1807945.  

It wouldn't be too much work, in my opinion, to intermittently send a false positive through the system for whatever the algorithm is supposed to detect (false RF triplets, for example), just to make sure this system works.
ID: 1807949 · Report as offensive
Mark StevensonProject Donor
Volunteer tester
Avatar

Send message
Joined: 8 Sep 11
Posts: 1574
Credit: 137,246,440
RAC: 92,002
United Kingdom
Message 1807964 - Posted: 8 Aug 2016, 16:26:49 UTC - in response to Message 1807949.  

You could always try reading this

http://seticlassic.ssl.berkeley.edu/about_seti/about_seti_at_home_1.html

It's from one of the links on the front page with lots of other links about seti@home
Life is what you make of it :-)

When i'm good i'm very good , but when i'm bad i'm shi#eloads better ;-) In't I " buttercups " p.m.s.l at authoritie !!;-)
ID: 1807964 · Report as offensive
S\

Send message
Joined: 18 Oct 11
Posts: 28
Credit: 4,719,993
RAC: 0
United States
Message 1807971 - Posted: 8 Aug 2016, 17:13:03 UTC - in response to Message 1807964.  

There are also "test signals" that are intentionally injected into the system so the SETI@home team can confirm that the hardware and software is working properly at all points through the system.


Got the answer. Thank you :)
ID: 1807971 · Report as offensive
Profile Somebody who doesn't support SETI anymore

Send message
Joined: 29 Apr 07
Posts: 15
Credit: 949,350
RAC: 0
Canada
Message 1808420 - Posted: 11 Aug 2016, 1:34:57 UTC - in response to Message 1807971.  

The software is open source, so anybody can test it, if they want.
ID: 1808420 · Report as offensive
Profile William Rothamel
Avatar

Send message
Joined: 25 Oct 06
Posts: 3335
Credit: 1,322,163
RAC: 0
United States
Message 1808471 - Posted: 11 Aug 2016, 9:20:25 UTC - in response to Message 1807945.  

only "non-terrestrial" signals are kept for further examination.
____________


How do it know ? Does the signal require a certain doppler shift that coincides with the Earth's rotation?
ID: 1808471 · Report as offensive
Profile Julie
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 28 Oct 09
Posts: 33496
Credit: 12,399,195
RAC: 10,738
Belgium
Message 1811001 - Posted: 21 Aug 2016, 12:01:59 UTC - in response to Message 1807935.  

I have always assumed the brainiacs at Berkeley knew what they were doing when creating the software.

+100
rOZZ
Music
Pictures
ID: 1811001 · Report as offensive
rob smithProject Donor
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 15235
Credit: 253,188,634
RAC: 324,659
United Kingdom
Message 1811011 - Posted: 21 Aug 2016, 12:42:18 UTC

The speed of Earth's rotation is very small when compared to the "red shift" of the Solar system with respect to other stars.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1811011 · Report as offensive
Profile William Rothamel
Avatar

Send message
Joined: 25 Oct 06
Posts: 3335
Credit: 1,322,163
RAC: 0
United States
Message 1811040 - Posted: 21 Aug 2016, 15:10:09 UTC - in response to Message 1811011.  
Last modified: 21 Aug 2016, 15:12:06 UTC

Yes that's right but probably not relevant to the issue here. Depending on the Latitude of the Radio Telescope, The antenna is rotating at several hundred miles per hour relative to the Earth's axis. The signal if there is one would be a point in the sky which would smear into an arced line which for 2 seconds of data would be around 1800 feet long or so. So we would expect to see a gaussian -like (Normally Distributed) curve as it came on center beam to edge. If the source were rotating we would also see a doppler shift over this time period.

Don't really know what I am talking about but this seems like what should happen.
ID: 1811040 · Report as offensive

Message boards : SETI@home Science : How do we know this software works?


 
©2017 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.