p, number of fitted parameters

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

p, number of fitted parameters

Fred
Hello Karen

In the article http://arxiv.org/pdf/1701.04817v1 where you and your team present AIJ, you invoke p, the number of fitted parameters.
I wonder if p = 6 parameters : R_p/R*, a/R*, T_C, b, u_1 and u_2

I also wonder if the algorithm uses the \chiĀ² in a probabilistic way to compute a p-value, or only to perform the downhill method.

Kind regards

Fred
Reply | Threaded
Open this post in threaded view
|

Re: p, number of fitted parameters

karenacollins
Administrator
If all transit model parameters are enabled for fitting and no detrending is enabled, the number of fitted parameters is 7 (the ones you list plus baseline flux). For every transit model parameter that is fixed/locked, subtract 1. For every detrend parameter included, add 1. Chi2 is used as part of the NelderMeade method (sometimes referred to as downhill simplex or Amoeba) to select the "best fit" transit model. We do not current implement MCMC fitting.

Karen
Reply | Threaded
Open this post in threaded view
|

Re: p, number of fitted parameters

Fred
Thank you Karen