- labels: 1195832 --> Plug-ins
Implement AoV periodogram. See details of Fortran code here, attached.
Diff:
--- old +++ new @@ -1,3 +1 @@ -Implement AoV periodogram. See details of Fortran code here: - -https://sourceforge.net/tracker/?func=detail&aid=3005590&group\_id=263306&atid=1152052 +Implement AoV periodogram. See details of Fortran code here, attached.
Brad Walter asked this very good question in email on June 15 2014:
"David I have a question about ANOVA and AOV in VStar. The sample sized per
bin are different. What are you doing to correct for that. The "normal"
ANOVA formulas assume equal sample size. That is almost never the case in
astronomy. We use equal time duration bins but over long periods of data.
See the cautions near the bottom of the following:
http://en.wikipedia.org/wiki/Analysis_of_variance . When I took the
Statistics course to become acquainted with R I found another web site that
went into the corrections in detail but I can't find where I saved the
reference. I am just curious what method you use."
One possible implication is the need for some kind of date compensated AoV.
Note that Brad's comment above also applies to ANOVA in general.
Does Foster advocate a particular approach in his light curve analysis book?
See Brad Walter's bug report re: "dropping" results:
On Feb 15, Alex Schwarzenberg-Czerny emailed me to say:
Dear David,
I came accross your valid doubts what to do i VStar
when population of bins is deplated (poor phase coverage).
It is explicitly described in my paper 2006MNRAS.365..165S,
see the code listing in the end.
Wish you good luck,
Alex
Log in to post a comment.