Re: determination of a minimum curve

From: Rene Brun <Rene.Brun_at_cern.ch>
Date: Fri, 08 Dec 2006 11:32:50 +0100

Hi Eric,

There are two cases:

CASE 1

---

You can fit a parametric function f(x,y) to your data set n,x,y. Use a TF2* f2 class to model
your f(x,y) function and find the minimum in x,y via TF2::GetMinimumXY

CASE 2

---

You do not have a model function for your data set. You can use the class TMultiDimFit
to find an automatic parametrization of your data set in the form of monomials,
Chebischev polynomials or Legendre functions. TMultiDimFit can generate this function
(including the C++ code). Then you can use this function like in case 1.

Rene Brun

eric.houla_at_free.fr wrote:
> Dear all,
>
> I have 3D data (Ncontents depending on x and y), and i'd like to find the
> minimum curve (in x,y) for that.
> Could you indicate me some ROOT tools in order to do that ?
>
> Thank you!
>
Received on Fri Dec 08 2006 - 11:32:59 MET