Re: conf_map_counts and C statistics

From: Marc Audard <audard_at_email.domain.hidden>
Date: Thu, 20 Mar 2003 08:28:47 -0500
John-

Thanks for the reply. I am somehow now confused by
the output of a plot:

I use Dave's script to make Monte Carlo simulations
of an individual line. When using the chisqr statistics,
most of the 24 simulated best fit values (in a 2-D parameter
plane) fall within the 68% confidence range region for
2 degrees of freedom. Now with the Cash statistics, most
of the fit values fall outside the region. They still
are in the 95% region, however they are "concentrated"
in one area, not distributed over the whole confidence region.

I attach a plot. The important information is:

a) the solid red (68%), green (95%), and blue (99%) curves,
  which correspond to the confidence regions for 2 parameters

b) the cyan dots, which correspond to the best-fit values
   of simulated data (which use the best-fit value to the real
    data as model input).

Do you have a possible explanation?

Regards,

Marc

----
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

Received on Thu Mar 20 2003 - 08:42:44 EST

This archive was generated by hypermail 2.2.0 : Tue Nov 13 2012 - 08:45:39 EST