Giovanni Siragusa | 31 Jul 15:20 2014
Picon
Picon

RooFit: failure in fitting for qmu or qmuA

Dear Root experts,
I'm trying to extract an upper limit and I'm having some failures like 
the following:

AsymptoticCalculator::EvaluateNLL -  value = nan for poi fixed at = 
0.123485    fit time : Real time 0:00:27, CP time 27.610
[#0] PROGRESS:Eval --      OBSERVED DATA :  qmu   = nan condNLL = nan 
uncond 185173
[#0] ERROR:Minization -- AsymptoticCalculator:  failure in fitting for 
qmu or qmuA 0.123485 return a dummy result
[#0] PROGRESS:Eval -- P values for  mu_SIG =  0.123485
     CLs      = nan +/- 0
     CLb      = nan +/- 0
     CLsplusb = nan +/- 0

[#0] WARNING:Eval -- HypoTestInverter - Skip invalid result for  point 
mu_SIG = 0.123485

The problem is that the fit is failing just near the value, where I 
would expect to set the limit.
What can I do to debug and understand it?

Thanks,
G.

Salvador Marti | 30 Jul 12:09 2014
Picon

include external png into a graphic output file

Hi

I guess this is quite naive question. The point is: I want to include 
the logo of my experiment (mylogo.png) on top of the histograms I draw, 
then save everything into a graphics file (.png, .eps...)

Which is the simplest way of doing this?

Thanks
Salva

Karl Loewenich | 27 Jul 19:27 2014
Picon

qtroot on 64 bit windows

Dear ROOTERS,
with big help fom Valeri I was able to install and use successsfully QtROOT under WindowsXP and Windows Vista on several 32 bit computers.    Unfortunately in my department the 32 bit computers have been substituted by 64 bit machines running Windows 7.  I have the MSVC2010 32 bit version installed and tried the procedure for installing QtROOT with the 32 and the 64 bit version of cygwin.A first error occurs when the "./configure" command is issued, because the architecture cygwin_nt-6.3 x86_64 1.7.30 is not known. When I force the win32 architecture with ./configure win32 ....
installation goes somewhat further and stops with error determining ROOT_SRCDIR/ROOT_OBJDIR.
These can be set manually but shortly after compilation fails again. Someone succeded in installing QtRoot on a 64 bit windows machine?
Actually I'm not interested so much in a "64 bit ROOT" on windows, I would be happy to be able to compile a 32 bit version on the 64 bit machine. I understand that building 64 bit ROOT on windows has major difficulties since no 64 bit windows binaries are available for ROOT. 
 
Thanks in advance 
 
Karl  
Dr Ian Korir | 26 Jul 10:03 2014
Picon

Mlphiggs.C tutorial query

The number of params is entered as [4] but the input into the is only 3, can someone explain this? I expected params[3] in line 98.





































































































































Xxxxxxxxxxxxxxxxxend.
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the National Nuclear Regulator on +27 12 674 7100 or email : Support <at> nnr.co.za. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. If you are not the intended recipient, you are notified that disclosing, copying, distributing or taking any action in reliance on the contents of this information is strictly prohibited
Alexandros Attikis | 24 Jul 17:03 2014
Picon
Picon

Accessing nth result of Min$/Max$?

Hi ROOTers,

quick question. Is there a way to get the nth-result of a Min$/Max$ operation?

For example, we know that Max$(array of elements) returns the maximum  value of the 
elements (leading-value). What if I want to get the next-to-maximum value (sub-leading-value) or
next-to-next-to leading value etc.? 
Is there a way to do this? 

If not, does anyone know if there a way around this?

Cheers,
Alexandros
Antonio Augusto Alves Jr | 24 Jul 05:14 2014
Picon
Picon

Using RooStats::SPlot with weighted datasets

Dear ROOTERS,

Please, I would like to know if the class RooStats::SPlot is compatible
with weighted datasets ?

I would like to compute sWeights and yields for a efficiency corrected
data set, as described in the section 4.5, page 16 of the
http://arxiv.org/pdf/physics/0402083v3.pdf.

Is the class implemented in a such way to take in consideration the
event weights when calculating the sWeights ?

Thanks you!
A.A.

--

-- 
Queriam-me casado, fútil, quotidiano e tributável? Queriam-me o
contrário disto, o contrário de qualquer coisa? Se eu fosse outra
pessoa, fazia-lhes, a todos, a vontade. Assim, como sou, tenham
paciência! Vão para o diabo sem mim, Ou deixem-me ir sozinho para o
diabo! Alvaro de Campos.

Dr. Harinder Singh Bawa | 22 Jul 19:29 2014
Picon

Re: hadd large number of files

Hi Dr. Davies,

I was using hadd.C where inside that c-code, I loop over 1000 files. It runs fine but I dont see it merge all the files after analysing the target file. I observed it doesnt add many files.(Dont know how many).

BTW: Could you share how you loop over 2000 files in hadd command. I mean Are you using bash kind of script . Could you share your snippet?

Harinder



On Tue, Jul 22, 2014 at 2:03 AM, Dr P J Davies <paul.davies <at> york.ac.uk> wrote:
I have used the linux command line hadd, which is now provided with root, to successfully merge over 2000 files. What kind of weird results are you seeing?

All the best, Paul Davies 


-- 
Dr P J Davies, Nuclear Physics Group, Department of Physics, 
University of York, Heslington, York, Tel: +44 (0)1904 432245.


On 22 July 2014 at 02:18:21, Dr. Harinder Singh Bawa (harinder.singh.bawa <at> gmail.com) wrote:

Hello,

I understand that there had been an issue with hadd.C not taking large number of files. Could anybody point out what is the maximum number of files one can merge in one go?

I am not able to hadd 150 files since it is giving me weird results. 2-3 files are OK. I didnt check upto many can be merged..

Is there any well made script which can handle the issue and gives the best results if someone have 1000 directories each having 1 file and need to merge all in one go?

Many thanks..Any suggestion is appreciated.

--
Dr. Harinder Singh Bawa

                                          
[web][facebook][youtube][twitter]





--
Dr. Harinder Singh Bawa

                                          
[web][facebook][youtube][twitter]


Dr. Harinder Singh Bawa | 22 Jul 03:18 2014
Picon

hadd large number of files

Hello,

I understand that there had been an issue with hadd.C not taking large number of files. Could anybody point out what is the maximum number of files one can merge in one go?

I am not able to hadd 150 files since it is giving me weird results. 2-3 files are OK. I didnt check upto many can be merged..

Is there any well made script which can handle the issue and gives the best results if someone have 1000 directories each having 1 file and need to merge all in one go?

Many thanks..Any suggestion is appreciated.

--
Dr. Harinder Singh Bawa

                                          
[web][facebook][youtube][twitter]


Arthur E. Snyder | 19 Jul 21:44 2014
Picon

(unknown)


Hi

I'm trying to do 'createHistogram' from a pdf from the pdf itself. It has only 'AbsRooRealProx' references
to the 'AbsRooReal' variariable which are in turn to ref'd to 'RooRealVar' arguments when object is createdd.

The code (hh,cc) is below.

The problem is that I can't find out how to invoke 'createHistogram' with proxys.
In this line (where X is a proxy) it fails to find the function.

     return ((RooAbsReal*)this)->createHistogram("h1D",X.arg());

This appears to be because 'X.arg()' which I thought gave me access to the underlying 'AbsRooReal' returns
does not inherit from 'RooAbsRealLValue' expected by 'createHistogram'.

Evidently I'm not doing this correctly ...

It seems I could like do this by putting an ordinary reference to 'RooRealVar' in 'HistPdf', but it does seem
like I should be able to get there from the proxy somehow? The inheritance through 'RooAbsLValue' needed
for 'createHistogram' seems to be bypassed by 'RooAbsReal' ...

How do I get around this?

-Arthur

  /*****************************************************************************
  * Project: RooFit                                                           *
  *                                                                           *
  * This code was autogenerated by RooClassFactory                            *
  *****************************************************************************/

// Your description goes here...

#include "Riostream.h"

#include "HistPdf.hh" 
#include "RooAbsReal.h" 
#include "RooAbsCategory.h" 
#include <math.h> 
#include "TMath.h" 
#include "TString.h"
#include "TH2F.h"
#include <iostream>
#include "Math/Interpolator.h"

ClassImp(HistPdf) ;

//3D
HistPdf::HistPdf(const char *name, const char *title,
 		 RooAbsReal& _X,
 		 RooAbsReal& _Y,
 		 RooAbsReal& _Z,
 		 RooAbsReal& _XTrue):
   RooAbsPdf(name,title),
   X("X","X",this,_X),
   Y("Y","Y",this,_Y),
   Z("Z","Z",this,_Z),
   XTrue("XTrue","XTrue",this,_XTrue) { 
}

//2D
HistPdf::HistPdf(const char *name, const char *title,
 		 RooAbsReal& _X,
 		 RooAbsReal& _Y,
 		 RooAbsReal& _XTrue):
   RooAbsPdf(name,title),
   X("X","X",this,_X),
   Y("Y","Y",this,_Y),
   XTrue("XTrue","XTrue",this,_XTrue) { 
}

//1D
HistPdf::HistPdf(const char *name, const char *title,
 		 RooAbsReal& _X,
 		 RooAbsReal& _XTrue) :
   RooAbsPdf(name,title),
   X("X","X",this,_X),
   XTrue("XTrue","XTrue",this,_XTrue) { 
}

HistPdf::HistPdf(const HistPdf& other, const char* name) :
   RooAbsPdf(other,name),
   X("X",this,other.X),
   Y("Y",this,other.Y),
   Z("Z",this,other.Z),
   XTrue("XTrue",this,other.XTrue) { 
}

TH1* HistPdf::createHistogram() {

   if(Y.name()=="") {
     return ((RooAbsReal*)this)->createHistogram("h1D",X.arg());
   }
   else if(Z.name()=="") {
     //    return this->createHistogram("h2D",X.arg(),RooFit::YVar(Y.arg()));

   } else {
     //    return this->createHistogram
     //      ("h3d",X.arg(),RooFit::YVar(X.arg()),RooFit::ZVar(Z.arg()));
   }
}

/*****************************************************************************
  * Project: RooFit                                                           *
  *                                                                           *
  * Histogram in X,Y and possibley Z (i.e., up to 3D)                         *
  *****************************************************************************/

#ifndef HISTPDF
#define HISTPDF

#include "RooAbsPdf.h"
#include "RooRealProxy.h"
#include "RooCategoryProxy.h"
#include "RooAbsReal.h"
#include "RooAbsCategory.h"
#include "TString.h"

#include <vector>

class HistPdf : public RooAbsPdf {

public:

   // Default c'tor.  Needed for streaming
   HistPdf() {} ;

   // 3D c'tor
   HistPdf(const char *name, const char *title,
 	  RooAbsReal& _X,
 	  RooAbsReal& _Y,
 	  RooAbsReal& _Z,
 	  RooAbsReal& _XTrue
 	  );

   // 2D c'tor
   HistPdf(const char *name, const char *title,
 	  RooAbsReal& _X,
 	  RooAbsReal& _Y,
 	  RooAbsReal& _XTrue
 	  );

   // 1D c'tor
   HistPdf(const char *name, const char *title,
 	  RooAbsReal& _X,
 	  RooAbsReal& _XTrue
 	  );

   // Copy c'tor
   HistPdf(const HistPdf& other, const char* name=0) ;

   // D'tor.  All variables clean up after themselves
   virtual ~HistPdf() { }

protected:

   // Proxies for paramters
   RooRealProxy X ;
   RooRealProxy Y ;
   RooRealProxy Z ;
   RooRealProxy XTrue ;

   //create hist  1D, 2D, or 3D hist depending on what's available of X,Y,Z
   TH1* createHistogram();

    // This is inherited for RooAbsPdf, and evaluates the PDF for a single event

private:

   ClassDef(HistPdf,1)

};

#endif

A.E. Snyder, The Former Group C/Kipac        \!c*p?/
SLAC Mail Stop #29                          ((.   .))
Box 4349                                        |
Stanford, Ca, USA, 94309                      '\|/`
e-mail:snyder <at> slac.stanford.edu                 o
phone:650-926-2701                              _
http://www.slac.stanford.edu/~snyder          BaBar
FAX:707-313-0250                          Collaboration
                                                 &
                                            Fermi/GLAST

Arthur E. Snyder | 19 Jul 21:40 2014
Picon

How to create histogram from proxy

I'm trying to create a histogram with pdf that has a proxy to 
'RooAbsReal'. I'm having trouble figuring out how to do it.

The code is:

A.E. Snyder, The Former Group C/Kipac        \!c*p?/
SLAC Mail Stop #29                          ((.   .))
Box 4349                                        |
Stanford, Ca, USA, 94309                      '\|/`
e-mail:snyder <at> slac.stanford.edu                 o
phone:650-926-2701                              _
http://www.slac.stanford.edu/~snyder          BaBar
FAX:707-313-0250                          Collaboration
                                                 &
                                            Fermi/GLAST

Antonio Augusto Alves Jr | 19 Jul 04:00 2014
Picon
Picon

fit parameter errors in weighted extended maximum likelihood fits

Dear Rooters,

Please, where can I find  more detailed information about the following change in  ROOT v5-34-19:

RooFit
  • Fix the computation of fit parameter errors in weighted extended maximum likelihood fits

Thank you
A.A.
--
Queriam-me casado, fútil, quotidiano e tributável? Queriam-me o contrário disto, o contrário de qualquer coisa? Se eu fosse outra pessoa, fazia-lhes, a todos, a vontade. Assim, como sou, tenham paciência! Vão para o diabo sem mim, Ou deixem-me ir sozinho para o diabo! Alvaro de Campos.

Gmane