Re: Datset T: One more bug gone

[Doug Welch <welch@physics.mcmaster.ca>]

Andrew,

Nice work - those are quite beautiful RR Lyr lightcurves!

One thing that was described in the original Welch & Stetson
paper was a way to determine the variability index distribution
in the presence of unknown or unspecified systematic errors.
Simply put, you pick one of the residuals in the pair from
one star and the other from a second star of similar brightness
(and crowding, if you like). Any correlation *can't* be due
to the object, since it isn't one object!! You can define a
95 or 99% confidence interval for the variability index using 
the real photometry for actual non-variables stars in this manner.

Statistically, most of the time,
Doug

On Fri, 2 Mar 2001, Andrew Bennett wrote:

> I found and fixed the bug that was causing the
> bimodal standard deviations for the V magnitudes
> in yesterday's version.
> 
> Sorry to keep on posting changes - but I'm
> excited about this stuff. It won't go on: I was
> supposed to start pruning the vines last month.
> 
> As Michael Richmond pointed out, the error had to be the
> inclusion of grossly wrong magnitudes arising from
> our old and well known problem of splitting and
> joining. It was. 1st program, 3rd subroutine, 20th
> line examined: using wrong flag mask, thus including 
> instead of excluding stars recognised by the original 
> analysis program as being confused. Change 8 characters 
> and rerun.
> 
> The improvement is considerable:
> http://www3.ns.sympatico.ca/andrew.bennett/tass/datasett.html
> 
> I really believe that we are now finding variable stars
> with the MK IV!
> 
> The noise floor problem remains and is particularly
> bad for the I-data where the standard deviation increases
> for the brightest stars to values more than 10 times
> the error expected from the faint star scatter. We have
> to fix this one before anybody will be able to see
> planet transits!
> 
> Andrew Bennett, Avondale Vineyard
> 

==============================================================
 Douglas L Welch     | Office/voicemail (905) 525-9140 x23186  
 Physics & Astronomy | FAX              (905) 546-1252 
 McMaster University | 
 Hamilton, Ontario   | 
 Canada L8S 4M1      | E-mail       welch@physics.mcmaster.ca

    On research leave at LLNL (Oct 1, 2000 - Apr 30, 2001)
        LLNL Office/voicemail         (925) 424-2857
        LLNL FAX                      (925) 423-0238
     Continue to use welch@physics.mcmaster.ca for e-mail
==============================================================