Re: [ROOT] Fitting Parameter Limits....

From: Ray Fliller III (rfliller@bnl.gov)
Date: Thu Aug 29 2002 - 17:56:26 MEST


On Thu, 29 Aug 2002, Eddy Offermann wrote:

> Hi Ray,
> 
> Optimizing with a function like "([0]*x+[1])/[2]" is not a reasonable
> thing to do. The matrix of this problem will be singular,because
> of the correlation between [2], [0] and [1]. Better use [0]*x+[1].
> 
> I know this does not answer your question but it avoids it :-)

I was only using this as an example (it is this with other
nasiter terms!).  Though I could do this in my function, it would not help
as [0],[1], and [2] are all physically meaningful to me.  If I use your
suggestion, I loose the information that I seek.  

Unfortunately avoidance is not the best solution in this case, though I
agree that if one can avoid problems, they are easily solved! :-)

-- 

 RAINMAN
      "How much does a piece of gum cost Ray?"
      "About $100."
      "How much does a new car cost Ray?"
      "`Bout $100."
      "There goes working for NASA!"

==================================================================
Ray Fliller: rfliller@bnl.gov      Office Phone: (631)-344-6124 
C-A Accelerator Physics            Dept. Fax: (631)-344-5954
Building 911B
Brookhaven National Lab
Upton, NY 11973                    Office: Room 211

http://www.agsrhichome.bnl.gov/People/rfliller/
==================================================================



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:51:06 MET