New version of the WS program

[Patrick Wils <patrickwils@yahoo.com>]

I have made some modifications to the WS program (again):

- The program no longer reads the number of data points of a star from
each line.  Instead it just takes the number of lines that are present
in the file.  The September data set e.g. says the first star contains
13 data points, while there are only 11 in the file.  This could have
caused a lot of problems.  Thanks to Michael Sallman for pointing this
out.
- The value of the "mean square successive difference" is now expressed
as 1 - SSSD/2/Variance, with SSSD the sum of the square successive
differences.  The closer this value approaches 1, the more probable the
star is a short period variable.
- Outliers have also been removed from the calculation of the normal WS
statistic.  The number of outliers is provided as well (as a
consequence some of the output fields have changed position).
- Some nights show a lot of scatter (e.g. JD2452497 in the September
data set is such a night).  These nights are best removed from the
calculations.  Just make a file with lines containing the two JDs
between which data should be discarded, and provide the name of this
file as the "$bad" parameter.

Feedback is always welcome.

Patrick



__________________________________________________
Do you Yahoo!?
Yahoo! News - Today's headlines
http://news.yahoo.com

wsv33.pl