Logo ROOT  
Reference Guide
 
Loading...
Searching...
No Matches
TFumiliMinimizer.cxx
Go to the documentation of this file.
1// @(#)root/fumili:$Id$
2// Author: L. Moneta Wed Oct 25 16:28:55 2006
3
4/**********************************************************************
5 * *
6 * Copyright (c) 2006 LCG ROOT Math Team, CERN/PH-SFT *
7 * *
8 * *
9 **********************************************************************/
10
11// Implementation file for class TFumiliMinimizer
12
13#include "TFumiliMinimizer.h"
14#include "Math/IFunction.h"
15#include "Math/Util.h"
16#include "TError.h"
17
18#include "TFumili.h"
19
20#include <iostream>
21#include <cassert>
22#include <algorithm>
23#include <functional>
24
25
26// setting USE_FUMILI_FUNCTION will use the Derivatives provided by Fumili
27// instead of what proided in FitUtil::EvalChi2Residual
28// t.d.: use still standard Chi2 but replace model function
29// with a gradient function where gradient is computed by TFumili
30// since TFumili knows the step size can calculate it better
31// Derivative in Fumili are very fast (1 extra call for each parameter)
32// + 1 function evaluation
33//
34//#define USE_FUMILI_FUNCTION
35#ifdef USE_FUMILI_FUNCTION
36bool gUseFumiliFunction = true;
37//#include "FumiliFunction.h"
38// fit method function used in TFumiliMinimizer
39
42#include "Fit/Chi2FCN.h"
43#include "TF1.h"
44#include "TFumili.h"
45
46template<class MethodFunc>
47class FumiliFunction : public ROOT::Math::FitMethodFunction {
48
50
51public:
52 FumiliFunction(TFumili * fumili, const ROOT::Math::FitMethodFunction * func) :
53 ROOT::Math::FitMethodFunction(func->NDim(), func->NPoints() ),
54 fFumili(fumili),
55 fObjFunc(0)
56 {
57 fObjFunc = dynamic_cast<const MethodFunc *>(func);
58 assert(fObjFunc != 0);
59
60 // create TF1 class from model function
61 fModFunc = new TF1("modfunc",ROOT::Math::ParamFunctor( &fObjFunc->ModelFunction() ) );
62 fFumili->SetUserFunc(fModFunc);
63 }
64
65 ROOT::Math::FitMethodFunction::Type_t Type() const { return fObjFunc->Type(); }
66
67 FumiliFunction * Clone() const { return new FumiliFunction(fFumili, fObjFunc); }
68
69
70 // recalculate data element using Fumili stuff
71 double DataElement(const double * /*par */, unsigned int i, double * g, double *) const {
72
73 // parameter values are inside TFumili
74
75 // suppose type is bin likelihood
76 unsigned int npar = fObjFunc->NDim();
77 double y = 0;
78 double invError = 0;
79 const double *x = fObjFunc->Data().GetPoint(i,y,invError);
80 double fval = fFumili->EvalTFN(g,const_cast<double *>( x));
81 fFumili->Derivatives(g, const_cast<double *>( x));
82
84 double logPdf = y * ROOT::Math::Util::EvalLog( fval) - fval;
85 for (unsigned int k = 0; k < npar; ++k) {
86 g[k] *= ( y/fval - 1.) ;
87 }
88
89 return logPdf;
90 }
91 else if (fObjFunc->Type() == ROOT::Math::FitMethodFunction::kLeastSquare ) {
92 double resVal = (y-fval)*invError;
93 for (unsigned int k = 0; k < npar; ++k) {
94 g[k] *= -invError;
95 }
96 return resVal;
97 }
98
99 return 0;
100 }
101
102
103private:
104
105 double DoEval(const double *x ) const {
106 return (*fObjFunc)(x);
107 }
108
109 TFumili * fFumili;
110 const MethodFunc * fObjFunc;
111 TF1 * fModFunc;
112
113};
114#else
116#endif
117//______________________________________________________________________________
118//
119// TFumiliMinimizer class implementing the ROOT::Math::Minimizer interface using
120// TFumili.
121// This class is normally instantiates using the plug-in manager
122// (plug-in with name Fumili or TFumili)
123// In addition the user can choose the minimizer algorithm: Migrad (the default one), Simplex, or Minimize (combined Migrad + Simplex)
124//
125//__________________________________________________________________________________________
126
127// initialize the static instances
128
132
133
135
136
138 fDim(0),
139 fNFree(0),
140 fMinVal(0),
141 fEdm(-1),
142 fFumili(nullptr)
143{
144 // Constructor for TFumiliMinimier class
145
146 // construct with npar = 0 (by default a value of 25 is used in TFumili for allocating the arrays)
147#ifdef USE_STATIC_TMINUIT
148 // allocate here only the first time
149 if (fgFumili == 0) fgFumili = new TFumili(0);
151#else
152 if (fFumili) delete fFumili;
153 fFumili = new TFumili(0);
155#endif
156
157}
158
159
161{
162 // Destructor implementation.
163 if (fFumili) delete fFumili;
164}
165
167 Minimizer()
168{
169 // Implementation of copy constructor (it is private).
170}
171
173{
174 // Implementation of assignment operator (private)
175 if (this == &rhs) return *this; // time saving self-test
176 return *this;
177}
178
179
180
182 // Set the objective function to be minimized, by passing a function object implement the
183 // basic multi-dim Function interface. In this case the derivatives will be
184 // calculated by Fumili
185
186 // Here a TFumili instance is created since only at this point we know the number of parameters
187 // needed to create TFumili
188 fDim = func.NDim();
190
191 if(func.HasGradient()) {
192 // In this case the function derivatives are provided
193 // by the user via this interface and there not calculated by Fumili.
194
195 fDim = func.NDim();
197
198 // for Fumili the fit method function interface is required
199 const ROOT::Math::FitMethodGradFunction * fcnfunc = dynamic_cast<const ROOT::Math::FitMethodGradFunction *>(&func);
200 if (!fcnfunc) {
201 Error("SetFunction","Wrong Fit method function type used for Fumili");
202 return;
203 }
204 // assign to the static pointer (NO Thread safety here)
205 fgFunc = nullptr;
206 fgGradFunc = const_cast<ROOT::Math::FitMethodGradFunction *>(fcnfunc);
208
209 return;
210 }
211
212 // for Fumili the fit method function interface is required
213 const ROOT::Math::FitMethodFunction * fcnfunc = dynamic_cast<const ROOT::Math::FitMethodFunction *>(&func);
214 if (!fcnfunc) {
215 Error("SetFunction","Wrong Fit method function type used for Fumili");
216 return;
217 }
218 // assign to the static pointer (NO Thread safety here)
219 fgFunc = const_cast<ROOT::Math::FitMethodFunction *>(fcnfunc);
220 fgGradFunc = nullptr;
222
223#ifdef USE_FUMILI_FUNCTION
224 if (gUseFumiliFunction) {
229 }
230#endif
231
232}
233
234void TFumiliMinimizer::Fcn( int & , double * g , double & f, double * x , int /* iflag */) {
235 // implementation of FCN static function used internally by TFumili.
236 // Adapt IMultiGenFunction interface to TFumili FCN static function
237 f = TFumiliMinimizer::EvaluateFCN(const_cast<double*>(x),g);
238}
239
240// void TFumiliMinimizer::FcnGrad( int &, double * g, double & f, double * x , int iflag ) {
241// // implementation of FCN static function used internally by TFumili.
242// // Adapt IMultiGradFunction interface to TFumili FCN static function in the case of user
243// // provided gradient.
244// ROOT::Math::IMultiGradFunction * gFunc = dynamic_cast<ROOT::Math::IMultiGradFunction *> ( fgFunc);
245
246// assert(gFunc != 0);
247// f = gFunc->operator()(x);
248
249// // calculates also derivatives
250// if (iflag == 2) gFunc->Gradient(x,g);
251// }
252
253double TFumiliMinimizer::EvaluateFCN(const double * x, double * grad) {
254 // function called to evaluate the FCN at the value x
255 // calculates also the matrices of the second derivatives of the objective function needed by FUMILI
256
257
258 //typedef FumiliFCNAdapter::Function Function;
259
260
261
262 // reset
263// assert(grad.size() == npar);
264// grad.assign( npar, 0.0);
265// hess.assign( hess.size(), 0.0);
266
267 double sum = 0;
268 unsigned int ndata = 0;
269 unsigned int npar = 0;
270 if (fgFunc) {
271 ndata = fgFunc->NPoints();
272 npar = fgFunc->NDim();
274 }
275 else if (fgGradFunc) {
276 ndata = fgGradFunc->NPoints();
277 npar = fgGradFunc->NDim();
279 }
280
281 // eventually store this matrix as static member to optimize speed
282 std::vector<double> gf(npar);
283 std::vector<double> hess(npar*(npar+1)/2);
284 std::vector<double> h(npar*(npar+1)/2);
285
286 // reset gradients
287 for (unsigned int ipar = 0; ipar < npar; ++ipar)
288 grad[ipar] = 0;
289
290
291 //loop on the data points
292//#define DEBUG
293#ifdef DEBUG
294 std::cout << "=============================================";
295 std::cout << "par = ";
296 for (unsigned int ipar = 0; ipar < npar; ++ipar)
297 std::cout << x[ipar] << "\t";
298 std::cout << std::endl;
299 if (fgFunc) std::cout << "type " << fgFunc->Type() << std::endl;
300#endif
301
302
303 // assume for now least-square
304 // since TFumili does not use errordef we must divide chi2 by 2
307
308 double fval = 0;
309 for (unsigned int i = 0; i < ndata; ++i) {
310 // calculate data element and gradient
311 // DataElement returns (f-y)/s and gf is derivatives of model function multiplied by (-1/sigma)
312 if (gUseFumiliFunction) {
313 fval = fgFunc->DataElement( x, i, &gf[0]);
314 }
315 else {
316 if (fgFunc != nullptr)
317 fval = fgFunc->DataElement(x, i, gf.data(), h.data() );
318 else
319 fval = fgGradFunc->DataElement(x, i, gf.data(), h.data());
320 }
321
322 // t.b.d should protect for bad values of fval
323 sum += 0.5 * fval * fval; // neeedd to divide chi2 by 2
324
325 for (unsigned int j = 0; j < npar; ++j) {
326 grad[j] += fval * gf[j];
327 for (unsigned int k = j; k < npar; ++ k) {
328 int idx = j + k*(k+1)/2;
329 //hess[idx] += gf[j] * gf[k];
330 hess[idx] += 0.5 * h[idx]; // h is gradient of full residual (2 * gf[j] n* gf[k] )
331 }
332 }
333 }
334 }
337
338
339 double fval = 0;
340
341 //std::cout << "\t x " << x[0] << " " << x[1] << " " << x[2] << std::endl;
342
343 for (unsigned int i = 0; i < ndata; ++i) {
344
345 if (gUseFumiliFunction) {
346 fval = fgFunc->DataElement( x, i, &gf[0]);
347 }
348 else {
349 // calculate data element and gradient
350 if (fgFunc != nullptr)
351 fval = fgFunc->DataElement(x, i, &gf[0], h.data());
352 else
353 fval = fgGradFunc->DataElement(x, i, &gf[0], h.data());
354 }
355
356 sum += fval;
357
358 for (unsigned int j = 0; j < npar; ++j) {
359 grad[j] += gf[j];
360 for (unsigned int k = j; k < npar; ++ k) {
361 int idx = j + k*(k+1)/2;
362 hess[idx] += h[idx];
363 }
364 }
365 }
368
369 double fval = 0;
370
371 for (unsigned int i = 0; i < ndata; ++i) {
372
373 if (gUseFumiliFunction) {
374 fval = fgFunc->DataElement(x, i, &gf[0]);
375 } else {
376 // calculate data element and gradient
377 if (fgFunc != nullptr)
378 fval = fgFunc->DataElement(x, i, &gf[0]);
379 else
380 fval = fgGradFunc->DataElement(x, i, &gf[0]);
381 }
382 sum -= fval;
383
384 for (unsigned int j = 0; j < npar; ++j) {
385 double gfj = gf[j]; // / fval;
386 grad[j] -= gfj;
387 for (unsigned int k = j; k < npar; ++k) {
388 int idx = j + k * (k + 1) / 2;
389 hess[idx] += gfj * gf[k]; // / (fval );
390 }
391 }
392 }
393 } else {
394 Error("EvaluateFCN", " type of fit method is not supported, it must be chi2 or log-likelihood");
395 }
396
397 // now TFumili excludes fixed parameter in second-derivative matrix
398 // ned to get them using the static instance of TFumili
399 double * zmatrix = fgFumili->GetZ();
400 double * pl0 = fgFumili->GetPL0(); // parameter limits
401 assert(zmatrix != nullptr);
402 assert(pl0 != nullptr);
403 unsigned int k = 0;
404 unsigned int l = 0;
405 for (unsigned int i = 0; i < npar; ++i) {
406 for (unsigned int j = 0; j <= i; ++j) {
407 if (pl0[i] > 0 && pl0[j] > 0) { // only for non-fixed parameters
408 zmatrix[l++] = hess[k];
409 }
410 k++;
411 }
412 }
413
414#ifdef DEBUG
415 std::cout << "FCN value " << sum << " grad ";
416 for (unsigned int ipar = 0; ipar < npar; ++ipar)
417 std::cout << grad[ipar] << "\t";
418 std::cout << std::endl << std::endl;
419#endif
420
421
422 return sum;
423
424}
425
426
427
428bool TFumiliMinimizer::SetVariable(unsigned int ivar, const std::string & name, double val, double step) {
429 // set a free variable.
430 if (fFumili == nullptr) {
431 Error("SetVariableValue","invalid TFumili pointer. Set function first ");
432 return false;
433 }
434#ifdef DEBUG
435 std::cout << "set variable " << ivar << " " << name << " value " << val << " step " << step << std::endl;
436#endif
437
438 int ierr = fFumili->SetParameter(ivar , name.c_str(), val, step, 0., 0. );
439 if (ierr) {
440 Error("SetVariable","Error for parameter %d ",ivar);
441 return false;
442 }
443 return true;
444}
445
446bool TFumiliMinimizer::SetLimitedVariable(unsigned int ivar, const std::string & name, double val, double step, double lower, double upper) {
447 // set a limited variable.
448 if (fFumili == nullptr) {
449 Error("SetVariableValue","invalid TFumili pointer. Set function first ");
450 return false;
451 }
452#ifdef DEBUG
453 std::cout << "set limited variable " << ivar << " " << name << " value " << val << " step " << step << std::endl;
454#endif
455 int ierr = fFumili->SetParameter(ivar, name.c_str(), val, step, lower, upper );
456 if (ierr) {
457 Error("SetLimitedVariable","Error for parameter %d ",ivar);
458 return false;
459 }
460 return true;
461}
462#ifdef LATER
463bool Fumili2Minimizer::SetLowerLimitedVariable(unsigned int ivar , const std::string & name , double val , double step , double lower ) {
464 // add a lower bounded variable as a double bound one, using a very large number for the upper limit
465 double s = val-lower;
466 double upper = s*1.0E15;
467 if (s != 0) upper = 1.0E15;
468 return SetLimitedVariable(ivar, name, val, step, lower,upper);
469}
470#endif
471
472
473bool TFumiliMinimizer::SetFixedVariable(unsigned int ivar, const std::string & name, double val) {
474 // set a fixed variable.
475 if (fFumili == nullptr) {
476 Error("SetVariableValue","invalid TFumili pointer. Set function first ");
477 return false;
478 }
479
480
481 int ierr = fFumili->SetParameter(ivar, name.c_str(), val, 0., val, val );
482 fFumili->FixParameter(ivar);
483
484#ifdef DEBUG
485 std::cout << "Fix variable " << ivar << " " << name << " value " << std::endl;
486#endif
487
488 if (ierr) {
489 Error("SetFixedVariable","Error for parameter %d ",ivar);
490 return false;
491 }
492 return true;
493}
494
495bool TFumiliMinimizer::SetVariableValue(unsigned int ivar, double val) {
496 // set the variable value
497 if (fFumili == nullptr) {
498 Error("SetVariableValue","invalid TFumili pointer. Set function first ");
499 return false;
500 }
502 double oldval, verr, vlow, vhigh = 0;
503 int ierr = fFumili->GetParameter( ivar, &name[0], oldval, verr, vlow, vhigh);
504 if (ierr) {
505 Error("SetVariableValue","Error for parameter %d ",ivar);
506 return false;
507 }
508#ifdef DEBUG
509 std::cout << "set variable " << ivar << " " << name << " value "
510 << val << " step " << verr << std::endl;
511#endif
512
513 ierr = fFumili->SetParameter(ivar , name , val, verr, vlow, vhigh );
514 if (ierr) {
515 Error("SetVariableValue","Error for parameter %d ",ivar);
516 return false;
517 }
518 return true;
519}
520
522 // perform the minimization using the algorithm chosen previously by the user
523 // By default Migrad is used.
524 // Return true if the found minimum is valid and update internal cached values of
525 // minimum values, errors and covariance matrix.
526
527 if (fFumili == nullptr) {
528 Error("SetVariableValue","invalid TFumili pointer. Set function first ");
529 return false;
530 }
531
532 // need to set static instance to be used when calling FCN
534
535
536 double arglist[10];
537
538 // error cannot be set in TFumili (always the same)
539// arglist[0] = ErrorUp();
540// fFumili->ExecuteCommand("SET Err",arglist,1);
541
542 int printlevel = PrintLevel();
543 // not implemented in TFumili yet
544 //arglist[0] = printlevel - 1;
545 //fFumili->ExecuteCommand("SET PRINT",arglist,1,ierr);
546
547 // suppress warning in case Printlevel() == 0
548 if (printlevel == 0) fFumili->ExecuteCommand("SET NOW",arglist,0);
549 else fFumili->ExecuteCommand("SET WAR",arglist,0);
550
551
552 // minimize: use ExecuteCommand instead of Minimize to set tolerance and maxiter
553
554 arglist[0] = MaxFunctionCalls();
555 arglist[1] = Tolerance();
556
557 if (printlevel > 0)
558 std::cout << "Minimize using TFumili with tolerance = " << Tolerance()
559 << " max calls " << MaxFunctionCalls() << std::endl;
560
561 int iret = fFumili->ExecuteCommand("MIGRAD",arglist,2);
562 fStatus = iret;
563 //int iret = fgFumili->Minimize();
564
565 // Hesse and IMP not implemented
566// // run improved if needed
567// if (ierr == 0 && fType == ROOT::Fumili::kMigradImproved)
568// fFumili->mnexcm("IMPROVE",arglist,1,ierr);
569
570// // check if Hesse needs to be run
571// if (ierr == 0 && IsValidError() ) {
572// fFumili->mnexcm("HESSE",arglist,1,ierr);
573// }
574
575
576 int ntot;
577 int nfree;
578 double errdef = 0; // err def is not used by Fumili
579 fFumili->GetStats(fMinVal,fEdm,errdef,nfree,ntot);
580
581 if (printlevel > 0)
582 fFumili->PrintResults(printlevel,fMinVal);
583
584
585 assert (static_cast<unsigned int>(ntot) == fDim);
586 assert( nfree == fFumili->GetNumberFreeParameters() );
587 fNFree = nfree;
588
589
590 // get parameter values and correlation matrix
591 // fumili stores only lower part of diagonal matrix of the free parameters
592 fParams.resize( fDim);
593 fErrors.resize( fDim);
594 fCovar.resize(fDim*fDim);
595 const double * cv = fFumili->GetCovarianceMatrix();
596 unsigned int l = 0;
597 for (unsigned int i = 0; i < fDim; ++i) {
598 fParams[i] = fFumili->GetParameter( i );
599 fErrors[i] = fFumili->GetParError( i );
600
601 if ( !fFumili->IsFixed(i) ) {
602 for (unsigned int j = 0; j <=i ; ++j) {
603 if ( !fFumili->IsFixed(j) ) {
604 fCovar[i*fDim + j] = cv[l];
605 fCovar[j*fDim + i] = fCovar[i*fDim + j];
606 l++;
607 }
608 }
609 }
610 }
611
612 return (iret==0) ? true : false;
613}
614
615
616// } // end namespace Fit
617
618// } // end namespace ROOT
#define f(i)
Definition RSha256.hxx:104
#define g(i)
Definition RSha256.hxx:105
#define h(i)
Definition RSha256.hxx:106
#define ClassImp(name)
Definition Rtypes.h:377
void Error(const char *location, const char *msgfmt,...)
Use this function in case an error occurred.
Definition TError.cxx:185
bool gUseFumiliFunction
char name[80]
Definition TGX11.cxx:110
Chi2FCN class for binned fits using the least square methods.
Definition Chi2FCN.h:46
class evaluating the log likelihood for binned Poisson likelihood fits it is template to distinguish ...
FitMethodFunction class Interface for objective functions (like chi2 and likelihood used in the fit) ...
virtual Type_t Type() const
return the type of method, override if needed
Type_t
enumeration specifying the possible fit method types
virtual double DataElement(const double *x, unsigned int i, double *g=nullptr, double *h=nullptr, bool fullHessian=false) const =0
method returning the data i-th contribution to the fit objective function For example the residual fo...
virtual unsigned int NPoints() const
return the number of data points used in evaluating the function
unsigned int NDim() const override
Number of dimension (parameters) .
virtual void UpdateNCalls() const
update number of calls
Documentation for the abstract class IBaseFunctionMultiDim.
Definition IFunction.h:61
virtual unsigned int NDim() const =0
Retrieve the dimension of the function.
double Tolerance() const
absolute tolerance
Definition Minimizer.h:315
unsigned int MaxFunctionCalls() const
max number of function calls
Definition Minimizer.h:309
int fStatus
status of minimizer
Definition Minimizer.h:391
int PrintLevel() const
minimizer configuration parameters
Definition Minimizer.h:306
Param Functor class for Multidimensional functions.
1-Dim function class
Definition TF1.h:214
TFumiliMinimizer class: minimizer implementation based on TFumili.
static ROOT::Math::FitMethodFunction * fgFunc
bool SetFixedVariable(unsigned int, const std::string &, double) override
set fixed variable (override if minimizer supports them )
TFumiliMinimizer(int dummy=0)
Default constructor (an argument is needed by plug-in manager)
std::vector< double > fParams
bool SetLimitedVariable(unsigned int ivar, const std::string &name, double val, double step, double, double) override
set upper/lower limited variable (override if minimizer supports them )
static double EvaluateFCN(const double *x, double *g)
implementation of FCN for Fumili when user provided gradient is used
~TFumiliMinimizer() override
Destructor (no operations)
bool Minimize() override
method to perform the minimization
static ROOT::Math::FitMethodGradFunction * fgGradFunc
TFumiliMinimizer & operator=(const TFumiliMinimizer &rhs)
Assignment operator.
static TFumili * fgFumili
bool SetVariableValue(unsigned int ivar, double val) override
set the value of an existing variable
std::vector< double > fErrors
std::vector< double > fCovar
void SetFunction(const ROOT::Math::IMultiGenFunction &func) override
set the function to minimize
bool SetVariable(unsigned int ivar, const std::string &name, double val, double step) override
set free variable
static void Fcn(int &, double *, double &f, double *, int)
implementation of FCN for Fumili
Double_t GetParameter(Int_t ipar) const override
Return current value of parameter ipar.
Definition TFumili.cxx:835
Double_t GetParError(Int_t ipar) const override
Return error of parameter ipar.
Definition TFumili.cxx:826
Bool_t IsFixed(Int_t ipar) const override
Return kTRUE if parameter ipar is fixed, kFALSE otherwise)
Definition TFumili.cxx:1060
Int_t ExecuteCommand(const char *command, Double_t *args, Int_t nargs) override
Execute MINUIT commands.
Definition TFumili.cxx:391
void PrintResults(Int_t k, Double_t p) const override
Prints fit results.
Definition TFumili.cxx:1483
Int_t GetNumberFreeParameters() const override
Return the number of free parameters.
Definition TFumili.cxx:814
Double_t * GetCovarianceMatrix() const override
Return a pointer to the covariance matrix.
Definition TFumili.cxx:784
Int_t GetStats(Double_t &amin, Double_t &edm, Double_t &errdef, Int_t &nvpar, Int_t &nparx) const override
Return global fit parameters.
Definition TFumili.cxx:904
const char * GetParName(Int_t ipar) const override
Return name of parameter ipar.
Definition TFumili.cxx:872
void FixParameter(Int_t ipar) override
Fixes parameter number ipar.
Definition TFumili.cxx:774
Double_t * GetPL0() const
Definition TFumili.h:95
void SetParNumber(Int_t ParNum)
Definition TFumili.cxx:168
Double_t * GetZ() const
Definition TFumili.h:102
Int_t SetParameter(Int_t ipar, const char *parname, Double_t value, Double_t verr, Double_t vlow, Double_t vhigh) override
Sets for parameter number ipar initial parameter value, name parname, initial error verr and limits v...
Definition TFumili.cxx:1641
Basic string class.
Definition TString.h:139
virtual void SetFCN(void(*fcn)(Int_t &, Double_t *, Double_t &f, Double_t *, Int_t))
To set the address of the minimization objective function called by the native compiler (see function...
Double_t y[n]
Definition legend1.C:17
Double_t x[n]
Definition legend1.C:17
Namespace for new Math classes and functions.
T EvalLog(T x)
safe evaluation of log(x) with a protections against negative or zero argument to the log smooth line...
Definition Util.h:64
BasicFitMethodFunction< ROOT::Math::IMultiGenFunction > FitMethodFunction
Definition Fitter.h:43
This file contains a specialised ROOT message handler to test for diagnostic in unit tests.
TLine l
Definition textangle.C:4
static uint64_t sum(uint64_t i)
Definition Factory.cxx:2345