September Variables from TOM3

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

Below is a repeat of the message I sent out earlier today.  I have since
redone the search with a slightly looser limit on sigma - 1.7 vs 2.  This
got 586 candidates, some of which were real variables that the previous
search did not pick up.  It is a question of how much junk you are willing
to look at.  The earlier message had .png files attached for a few stars. 
I have deleted this attachment.
---

For fun I try to look for variables. A good time to do this is before I
send in the monthly data. That way I have the first crack at the fun. There
is now enough data that one can be pretty sure about a possible variable. 
 
This month I have collected all the TOM3 data and searched it for
variables. 
 
The search:
 
1) WS > 10
2) 20 or more points
3) After eliminating the three largest and smallest V data points, the
sigma of the remainder is more than twice the expected sigma for noise at
the given magnitude.
 
This produced 229 stars. I then looked these up in Vizier using the GCVS
and Suspected Variables list. 51 matches were found using a 20 arc second
search radius. My guess is that most of the stars on this list are really
variables. There are a few that might be just noise, but that is
statistics, and some noise always gets through if you want to get the good
stuff. 
 
Two lists are attached:
 
coords The list of stars in a format suitable for insertion into VizieR.
coords.ckd The same list with the found variable names appended - may cause
trouble reading into VizieR
 
I have not attached the first 10 .png files as a "random" sample of the
data as it may have kept the message from getting out..
 
I have data files and .png files for all these stars but the total is a few
MB. I will send these files to anyone that asks for them, or send them to
someone who wants to put them up on a site. 
 
There seem to be more short period and fewer long period variables than I
expected. Perhaps I need to do more work on the search algorithm. There is
enough data now so that those of you with such a bent could start
developing search tools. 
 
Happy data munching.
 
Tom Droege

coords

coords.ckd