Message boards :
Number crunching :
how to view a wu?
Message board moderation
Author | Message |
---|---|
PhonAcq Send message Joined: 14 Apr 01 Posts: 1656 Credit: 30,658,217 RAC: 1 |
Is there a simple way (without unwinding the seti code) to view the decoaded contents of each wu? I'd like to see an example of the raw data. Perhaps this would be good to include in wiki. May this Farce be with You |
Janus Send message Joined: 4 Dec 01 Posts: 376 Credit: 967,976 RAC: 0 |
The results or the input? The results are pure XML text, so you can read that with any ordinary text browser. The input is a combination of descriptive XML and a small sample of radiotelescope data. |
PhonAcq Send message Joined: 14 Apr 01 Posts: 1656 Credit: 30,658,217 RAC: 1 |
I'm interested in seeing the input. The actual numbers that are processed are coded for download. I'd like to see what a typical example of the raw (input) data looks like. I guess it would be keen to be able to have a viewer that allows each client to view the wu's before processing and correlate what is seen to what is computed, although it may only be helpful to see corrupted wu's I suppose. May this Farce be with You |
MattDavis Send message Joined: 11 Nov 99 Posts: 919 Credit: 934,161 RAC: 0 |
I'm interested in seeing the input. The actual numbers that are processed are coded for download. I'd like to see what a typical example of the raw (input) data looks like. 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 01110110001011010010101010101110110001011010010101010101110110001011010010101 ----- |
PhonAcq Send message Joined: 14 Apr 01 Posts: 1656 Credit: 30,658,217 RAC: 1 |
Is this an attempt at humor or are you sharing your limited genome? I'm interested in seeing the input. The actual numbers that are processed are coded for download. I'd like to see what a typical example of the raw (input) data looks like. May this Farce be with You |
Tetsuji Maverick Rai Send message Joined: 25 Apr 99 Posts: 518 Credit: 90,863 RAC: 0 |
Is this an attempt at humor or are you sharing your limited genome? It's not a joke. The data in the WU are represented in 1bit. But inside it's represented as 1.0 and -1.0 in single float. I made a dumping program. If you like I will upload. However the raw data are acquired in 2 bits, then splitted in many frequency ranges, and delivered in each frequency band in 1 bit. They are mentioned in the head of a WU. For further details see in this page Luckiest in the world. WMD = Weapon of Mass Distraction. Click this table. |
PhonAcq Send message Joined: 14 Apr 01 Posts: 1656 Credit: 30,658,217 RAC: 1 |
Thanks, I've read that page but don't understand it. Your upload may help. I infer the spectrum analysis process yields into each wu the power in a narrow 10KHz band as a function of time for about 100 sec. The power is recorded as a one or zero (although you indicate it takes 2 bits to store this info). So there must be some sort of noise threshold (in mW) to define a zero and another one (in mW) to define a one. If my interpretation is correct, how are the thresholds set? Sorry to be so dense. May this Farce be with You |
Tetsuji Maverick Rai Send message Joined: 25 Apr 99 Posts: 518 Credit: 90,863 RAC: 0 |
okay, here's the workunitdump.exe. You put this and a workunit named "work_unit.sah" and launch workunitdump.exe. Then you'll find some other files. The raw data are in one of them, "workdata.txt" workunitdump.zip but be careful, they are complex values. So real part and imaginary part are listed in turns. There are 2097152 numbers in workdata.txt, but actually data of 1048576 points. BTW now I'm up to the ears in SETI beta test project, but it lacks morals and is messy...there have been 2 versions of SETI@home enhanced clients released, out of which only the newer one must be used now, but some people are still using the old optimized one, causing mixture of different results for the same WUs. Damn it!! And that, now it takes 10 to 30 (or 50?) hours to crunch a WU!! Crazy! Luckiest in the world. WMD = Weapon of Mass Distraction. Click this table. |
PhonAcq Send message Joined: 14 Apr 01 Posts: 1656 Credit: 30,658,217 RAC: 1 |
seti-beta: is there a mathematical reason the wu's can't be broken down to run in less time? or is it a network traffic thing, or something else? I ran your program and got the following error: Can't open init data file - running in standalone mode Is this a normal exit? Oh yeah. Why is the raw data complex valued? I understand the FFT would be but why is the raw? (You see I think I am missing a step in the analysis.) Oh, yeah^2. Is the source code for your applet available? I'd like to learn a bit there too. May this Farce be with You |
Tetsuji Maverick Rai Send message Joined: 25 Apr 99 Posts: 518 Credit: 90,863 RAC: 0 |
seti-beta: is there a mathematical reason the wu's can't be broken down to run in less time? or is it a network traffic thing, or something else? oh...I don't know. The data may be cut in halves (53.5seconds instead of 107 seconds.) But...think of CPDN. I myself haven't tried it, but it is said to take several weeks to crunch a WU. I cannot do anything....sorry Recently devs made some "BAD" WUs, with wrong headers which demand almost infinite chirp/fft pairs, make seti_enhanced client consume almost infinite memory, eating swap, and the cruncher couldn't begin math. I reported it to Eric and it was fixed soon......beta project is messy.
yes. It's just a modified version of setiathome client spitting data and exiting before calculation :) So the behavior is similar to the seti client. Easy (lazy?) job ;)
I mentioned "raw", but actually it is not raw data, because real raw data acquired by the radar detector in Arecibo, which is real value, is divided (splitted) by time as well as by frequency (107 seconds, 10khz width) by splitter, which uses FFT producing complex results for WUs (yeah, it uses fftw.) So "raw" data in a WU is in 1bit complex format. Luckiest in the world. WMD = Weapon of Mass Distraction. Click this table. |
©2024 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.