Hi Francois-Xavier, You have an array n x[n],y[n],z[n] with a sharp exponential distribution. You fill a binned histogram with your non-equidistant points. During the fit, the center of the bin will be taken to evaluate the function. In your case, this is not correct. You can see the effect by gradually increasing the number of bins of your histogram. The solution to this problem would be a TGraph3D class preserving the original coordinates. Rene Brun GENTIT Francois-Xavier DAPNIA wrote: > > Dear rooters, > In the following small CINT macro lemaire1.C I make a 2d fit and calculate > after that the chi2 of the fit. The problem is that the chi2 I calculate is > not equal to the chi2 published by ROOT (which in fact is calculated by > Minuit). Have you any idea of what happens? > The macro is simply launched by: > > .x lemaire1.C > > Thanks a lot for your help > > François-Xavier Gentit > DAPNIA/SPP CEA Saclay > http://gentit.home.cern.ch/gentit/ > > <<lemaire1.C>> > > -------------------------------------------------------------------------------- > > lemaire1.CName: lemaire1.C > Type: unspecified type (application/octet-stream)
This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:09 MET