MaCh3Factory.h File Reference

Factory methods for MaCh3 software which streamline initialisation of different objects. More...

#include "Fitters/FitterBase.h"
#include "Fitters/MR2T2.h"
#include "Fitters/DelayedMR2T2.h"
#include "Fitters/PSO.h"
#include "Fitters/LikelihoodFit.h"
#include "Parameters/ParameterHandlerGeneric.h"

Include dependency graph for MaCh3Factory.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Functions
template<typename CovType >
std::unique_ptr< CovType >	MaCh3CovarianceFactory (manager *FitManager, const std::string &PreFix)
	MaCh3 Factory initiates one of implemented fitting algorithms. More...
template<typename SampleType >
std::vector< SampleType * >	MaCh3SampleHandlerFactory (const std::vector< std::string > &SampleConfig, ParameterHandlerGeneric *xsec)
	Factory function for creating SampleHandler and initialisation with systematic. More...

Detailed Description

Factory methods for MaCh3 software which streamline initialisation of different objects.

Author

Kamil Skwarczynski

Definition in file MaCh3Factory.h.

Function Documentation

MaCh3CovarianceFactory()

template<typename CovType >

std::unique_ptr<CovType> MaCh3CovarianceFactory	(	manager *	FitManager,
		const std::string &	PreFix
	)

MaCh3 Factory initiates one of implemented fitting algorithms.

Parameters

FitManager

pointer to Manager class

Note

Example YAML configuration:

    General:
      FittingAlgorithm: ["MCMC"]
std::unique_ptr<FitterBase> MaCh3FitterFactory(manager *FitManager);
 
    @brief Initializes the config manager class and allows overriding settings via command-line arguments.
    @param argc number of arguments
    @param argv name of arguments
    @return A unique pointer to the initialized `manager` instance with optional overrides applied.
    @example Usage examples:
    ```
    ./bin/MCMCTutorial Inputs/FitterConfig.yaml General:OutputFile:blarb.root
    ./bin/MCMCTutorial Inputs/FitterConfig.yaml General:OutputFile:blarb.root General:MCMC:NSteps:50000
    ```
std::unique_ptr<manager> MaCh3ManagerFactory(int argc, char **argv);
 
// ********************************************
    @brief Factory function for creating a covariance class for systematic handling.
   
    @param FitManager Pointer to the manager class that holds the configuration settings.
    @param PreFix Prefix, for example Xsec, then code will look for XsecCovFile
    @return Pointer to the initialized covarianceXsec matrix object.
   
    @note Example YAML configuration:
    @code
   MaCh3CovarianceFactory(, Xsec)
   
    General:
      Systematics:
          XsecCovName: "xsec_cov"
          XsecCovFile: ["inputs/blarb1.yaml",
                        "inputs/blarb2.yaml"]
          XsecFix: ["Param_0",
                    "Param_1"]
          XsecPCAThreshold: -1
          #XsecPCAThreshold: 0.00001
   
          XsecPCAParams: [-999, -999]
          XsecStepScale: 0.0075

Todo:

add adaptive stuff

Definition at line 67 of file MaCh3Factory.h.

                                                                                           {
// ********************************************
  // config for our matrix
  YAML::Node Settings = FitManager->raw()["General"]["Systematics"];
  auto CovMatrixName = Get<std::string>(Settings[std::string(PreFix) + "CovName"], __FILE__, __LINE__);
  MACH3LOG_INFO("Initialising {} matrix", CovMatrixName);
 
  // yaml files initialising out matrix
  auto CovMatrixFile = Get<std::vector<std::string>>(Settings[std::string(PreFix) + "CovFile"], __FILE__, __LINE__);
 
  // PCA threshold, -1 means no pca
  auto PCAThreshold = GetFromManager<int>(Settings[std::string(PreFix) + "PCAThreshold"], -1);
  // do we pca whole matrix or only submatrix
  auto PCAParamRegion = GetFromManager<std::vector<int>>(Settings[std::string(PreFix) + "PCAParams"], {-999, -999});
 
  auto CovObject = std::make_unique<CovType>(CovMatrixFile, CovMatrixName, PCAThreshold, PCAParamRegion[0], PCAParamRegion[1]);
 
  // Fill the parameter values with their nominal values
  // should _ALWAYS_ be done before overriding with fix or flat
  CovObject->SetParameters();
 
  auto FixParams = GetFromManager<std::vector<std::string>>(Settings[std::string(PreFix) + "Fix"], {});
 
  // Fixed CovObject parameters loop
  if (FixParams.size() == 1 && FixParams.at(0) == "All") {
    for (int j = 0; j < CovObject->GetNumParams(); j++) {
      CovObject->ToggleFixParameter(j);
    }
  } else {
    for (unsigned int j = 0; j < FixParams.size(); j++) {
      CovObject->ToggleFixParameter(FixParams.at(j));
    }
  }
  //Global step scale for matrix
  auto StepScale = Get<double>(Settings[std::string(PreFix) + "StepScale"], __FILE__, __LINE__);
 
  CovObject->SetStepScale(StepScale);
 
  // Adaptive MCMC stuff
  if (CheckNodeExists(FitManager->raw(), "AdaptionOptions")) {
    CovObject->InitialiseAdaption(FitManager->raw());
  }
  
  return CovObject;
}

MaCh3SampleHandlerFactory()

template<typename SampleType >

std::vector<SampleType*> MaCh3SampleHandlerFactory	(	const std::vector< std::string > &	SampleConfig,
		ParameterHandlerGeneric *	xsec
	)

Factory function for creating SampleHandler and initialisation with systematic.

Template Parameters

SampleType

The class type of the sample to create, e.g., SampleHandlerTutorial.

Parameters

SampleConfig	Path to sample config.
xsec	A pointer to a ParameterHandlerGeneric object for cross-section systematic settings.

Returns

Vector of SampleType object, initialized and ready for use.

Note

Example

auto mySamples = MaCh3SampleHandlerFactory<SampleHandlerTutorial>(SampleConfig, xsec);

Definition at line 126 of file MaCh3Factory.h.

                                                                              {
// ********************************************
  std::vector<SampleType*> Handlers(SampleConfig.size());
  for (size_t i = 0; i < SampleConfig.size(); ++i)
  {
    // Instantiate the sample using the specified class type
    SampleType* Sample = new SampleType(SampleConfig[i], xsec);
    Sample->Reweight();
 
    // Obtain sample name and create a TString version for histogram naming
    std::string name = Sample->GetTitle();
    TString NameTString = TString(name.c_str());
 
    // Clone the 1D histogram with a modified name
    if (Sample->GetNDim() == 1) {
      auto hist = static_cast<TH1D*>(Sample->GetMCHist(1));
      Sample->AddData(hist);
    } else {
      auto hist = static_cast<TH2D*>(Sample->GetMCHist(2));
      Sample->AddData(hist);
    }
    Handlers[i] = Sample;
  }
  return Handlers;
}