Re: question about Gregory-Loredo algorithm in SITAR

On Feb 29, 2012, at 1:36 PM, juan luna wrote:

> I would like to know if the parameter "p" that is in the output of the sitar_glvary() function is defined as the Variability index or the Probability of variable signal used by GLVARY tool in CIAO. Values of ~0.5 means no variability?

The sitar_glvary function is an independent version of the CIAO glvary tool.  It should give you results pretty close to the CIAO tool, but not necessarily exactly the same. (Hopefully any differences are mostly "in the noise", so to speak.)

But, yes, "p" is the probability that a variable lightcurve is a better description.  So, 0.5 means no *detected* variability.  A constant is as equally good a description as any of the possible evenly binned lightcurves that were tried.  Which of course, could just be a function of S/N.  

Also, note that it does not check unevenly sampled lightcurves.

I haven't spent tons and tons of time calibrating G-L.  Although you can take a look at the Chandra source catalog characterization paper (Primini et al. 2011, ApJ Sup, 194, p.  37), specifically Fig. 44.  That gives you an idea of how G-L stacks up as regards white noise.  Overall, it's a little pessimistic in finding variability, but it is looking for a specific kind of variability, i.e., things that are well-described by an evenly spaced lightcurve.

-Mike



----
You received this message because you are
subscribed to the isis-users list.
To unsubscribe, send a message to
isis-users-request_at_email.domain.hiddenwith the first line of the message as:
unsubscribe