Re: [ROOT] busy flag problem

From: Rene Brun (Rene.Brun@cern.ch)
Date: Tue Dec 03 2002 - 15:02:03 MET


Hi Avdhesh,

Please could you identify which version of ROOT you are using?
In case you use a version older than 3.03/09, please try 3.03/09 first.
If the symptom persists, send me a small but complete script reproducing
the problem.

Rene Brun

On Tue, 3 Dec 2002, Avdhesh Chandra wrote:

> 
> Hi,
>       I have arrays of histogram define like
>   TH1F *Dca[8];
>   Dca[0] = new TH1F("Dca_0"," ",24000,-1.2,1.2);
>   Dca[1] = new TH1F("Dca_1"," ",24000,-1.2,1.2);
>    ......    .......
> 
>     In my code when i am filling these histograms i am facing no problem,
> but when i am trying to fit these histograms like
> 
>   for(Int_t i=0; i<8; i++)
>     { gausfit(Dca[i],sigma,errsigma,chisq,ndf);
>       Sigdca  [i] = sigma ;         Errsigdca[i] = errsigma ;
>       Chisq_1a[i] = chisq;          Ndf_1a[i] = ndf ;
>     }
> 
>           where gausfit is defined like
>    void gausfit( TH1F *histname,Double_t &sigma, Double_t &errsigma,
> 	                        Double_t &chisq, Double_t &ndf)
> 
>    I am getting following error message
> --------------------------------------------------------
>  FCN=19.9171 FROM MIGRAD    STATUS=CONVERGED     248 CALLS         249
> TOTAL
>                      EDM=2.10802e-09    STRATEGY= 1      ERROR MATRIX
> ACCURATE
>   EXT PARAMETER                                   STEP         FIRST
>   NO.   NAME      VALUE            ERROR          SIZE      DERIVATIVE
>    1  Constant     1.06828e+00   6.35883e-02   1.35696e-04   8.02646e-04
>    2  Mean         4.51991e-02   9.17302e-01   1.96188e-03  -5.30292e-05
>    3  Sigma        1.55634e+00   1.40853e+00   9.03530e-03** at limit **
> <TCanvas::MakeDefCanvas>: created default TCanvas with name c1
> 
>  *** Break *** segmentation violation
> Root > Function gausfit() busy flag cleared
> Function Loop() busy flag cleared
> -----------------------------------------------------------------
> 
>   This fit function works fine if I use different historgams
>                        (not arrays of historgams)
> 
>   Please let me know what should be the correct way of doing this.
> 
> Thanks
> Avdhesh
> 
> 



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:51:22 MET