Twin Star Problems

["Thomas Droege" <tdroege2@earthlink.net>]

I have been working on matching the tass star list to Brian Skiff's big
photometry file.  When I did, I found lots of problems with the "twin"
stars.  As most of you know, the program I have written to collect the star
measurements together is not perfect.  Lots of  "twin" stars get through. 
That is the data points for a star will be grouped in two or more groupings
each reported as a different star.  I have long determined to write a
second pass program for collect_stars.pl which would fix this problem.  I
have finally done so.  

I was surprised at the result.  One list of 30 million measurements found 3
millions stars with collect_stars.pl but with the second pass this was
reduced to 2.4 million.  So there are 20% twins.  Wow!  I did not think it
was that bad.  I still have to do more checking because when I put 30
millions stars into the program only 29,999,991 came out.  (I think I have
found the problem)  Programming is a lot like accounting, when you see a
penny error you had better look until it is found because it might really
be a million dollar error.  

I quickly plotted the result to see if the twin stars caused the error
distribution on a mag-mag sigma plot to be bad.  Well, there is always
hope, but there was no obvious improvement.  

All this software work should lead to some new variable star hunting lists
in the near future.  


Tom  Droege Jennifer Malpass
tdroege2@earthlink.net