Data Set

[Tom Droege <tdroege2@earthlink.net>]

A big batch of Data Set 23 went out today.  More will go out 
tomorrow.  There is still time to get in on the great variable star 
hunt.  We could have a contest, but in the past there has been no response 
to such.  So how about posting your candidates?

OK, my reason for sending out these disks is to encourage the development 
of automated techniques to find variable stars in the data.  Welch_Stetson 
is a start.  Doug has optimized the algorithm for our data.  But I want to 
look at each star to confirm that it really looks like it is 
variable.  With some "confidence" number that goes along with it.

You will see that the plotting routine needs some work when there is more 
than one day of observation.  Rich is working on a better scheme.   I 
should point out that I have fudged some of the numbers in Rich's code to 
try to get the plots more spread out in y.  So if it does funny things, it 
may not be his fault as I may not have understood the change that I made.

There is also the problem of finding the longer periods.  I did not find 
any.  Well, at least any that seemed to have enough statistics that I could 
be sure.

I think there is a lot of adventure to be had searching this data.  I can 
assure you that you will look at a lot of plots before you find the first 
convincing variable star.  A hint, I did not find any in the first few 
thousand stars, then there was a bunch.

If you stick to the collected.all list we can all compare stars by their 
number on that list.  You will want to look at each .cal file separately, I 
think.  So this will not always be possible.  RA and Dec will always get us 
to agree on a star.

Still time to get a copy of DS-23.

Happy hunting.

Tom Droege