Likelihood Class Reference

A likelihood object, consisting of parameters. More...

#include <XeStat.h>

Inheritance diagram for Likelihood:
XeCombinable XeStat XeMath XeObject XeCore XeCore LikelihoodFromDataSet ProfileLikelihood ebLKTest CombinedProfileLikelihood PLcountingSB S1S2PL

List of all members.

Public Member Functions

void activateParameter (LKParameter *param, bool active=true)
 Activate (add or remove) a parameter with its own id.
void addParameter (LKParameter *param, int id=SAME)
 add an existing parameter
void addParameter (int id, int type, string nam, double initialVal, double step, double mi, double ma)
 Add a parameter.
void addParameterTolerant (LKParameter *param)
 Add a parameter with its own id, wheter already added or not.
bool checkConsistency ()
void clearTheParameters ()
virtual double computeTheLogLikelihood ()=0
void forceNParametersOfInterest (int nF)
int getNActiveParameters ()
int getNMinuitParameters ()
int getNNuisanceParameters ()
int getNParameters (int type)
int getNParametersForChi2 ()
int getNParametersOfInterest ()
int getNTotalParameters ()
LKParametergetParameter (int id)
map< int, LKParameter * > * getParameters ()
double getParameterValue (int id)
void ignoreParameter (int id)
bool inCombinedMode ()
 Likelihood ()
 Empty constructor for root.
 Likelihood (string name)
 Constructor.
void listParameters ()
int mapMinuitParameters (bool freezeParametersOfInterest)
double maximize (bool freezeParametersOfInterest)
bool parameterExists (int p)
 Check wheher a parameter exists.
void printCurrentHeader ()
void printCurrentParameters ()
void printInitialHeader ()
void printInitialParameters ()
void printResultParameters ()
void removeParameter (int id, bool tolerant=false)
 Remove a parameter giving its id.
void removeParameter (LKParameter *param, bool tolerant=false)
 Remove a parameter with its own id.
void replaceParameter (LKParameter *param, int id=SAME)
 Replace a parameter.
void resetParameters ()
void setCombinedMode (bool mode=true)
void setCurrentValues (const double *v, const double *ers=NULL)
void setCurrentValuesInMinuitUnits (const double *v, const double *e=NULL)
void setInitialValue (int id, double v)
void setParameterType (int id, int type)
void setParameterValue (int id, double v)
void setTheExperiment ()
 pass the experiment number to whoever needs it after it was set.

Static Public Member Functions

static double shapeLikelihood (vector< int > *bins, int nDist, double **dists, double *norm)
 return the shape LL that a set of values comes from a mixture of tabulated parent distribution.
static double shapeLikelihood (vector< double > *values, vector< XeDist * > &dists, vector< double > &weights)
 return the shape LL that a set of values comes from a mixture of parent distributions.
static double shapeLikelihood (vector< double > *values, XeDist *dist)
 return the shape LL that a set of values comes from a parent distribution

Protected Member Functions

bool checkParameter (int p, bool shouldExist)
void clear ()
void setup ()
 Technical routine for constructors.

Protected Attributes

bool combinedMode
int currentId
int forcedNParametersOfInterest
vector< LKParameter * > MinuitParameters
int nActiveParameters
int nNuisanceParameters
int nParametersOfInterest
map< int, LKParameter * > parameters

Detailed Description

A likelihood object, consisting of parameters.

This is a virtual class


Constructor & Destructor Documentation

Likelihood::Likelihood ( string  name  ) 

Constructor.

Parameters:
name name of the object

Member Function Documentation

void Likelihood::activateParameter ( LKParameter param,
bool  active = true 
)

Activate (add or remove) a parameter with its own id.

Parameters:
param pointer to existing LKParameter
active activate(true) or disable (true)
void Likelihood::addParameter ( LKParameter param,
int  id = SAME 
)

add an existing parameter

Parameters:
param pointer to existing LKParameter
id requested id, or SAME (the one coded in the parameter), or AUTO
void Likelihood::addParameter ( int  id,
int  type,
string  nam,
double  initialVal,
double  step,
double  mi,
double  ma 
)

Add a parameter.

Parameters:
id either its hard coded id or the keyword AUTO
type NUISANCE_PARAMETER,PARAMETER_OF_INTEREST or FIXED_PARAMETER
name parameter name
initialVal initial value
step step for minimisation
mi minimum value
ma maximum value
void Likelihood::addParameterTolerant ( LKParameter param  ) 

Add a parameter with its own id, wheter already added or not.

Parameters:
param pointer to existing LKParameter
virtual double Likelihood::computeTheLogLikelihood (  )  [pure virtual]
Returns:
Pure virtual method returning the log likelihood
Parameters:
none All parameters values set thru LKParameter

Implemented in S1S2PL, LikelihoodFromDataSet, CombinedProfileLikelihood, and PLcountingSB.

void Likelihood::removeParameter ( int  id,
bool  tolerant = false 
)

Remove a parameter giving its id.

Parameters:
id identifier
tolerant don't issue a warning if it does not exist
void Likelihood::removeParameter ( LKParameter param,
bool  tolerant = false 
)

Remove a parameter with its own id.

Parameters:
param pointer to existing LKParameter
tolerant don't issue a warning if it does not exist
void Likelihood::replaceParameter ( LKParameter param,
int  id = SAME 
)

Replace a parameter.

If exists, delete the existing one

Parameters:
param pointer to existing LKParameter
id requested id, or SAME (the one coded in the parameter)
static double Likelihood::shapeLikelihood ( vector< int > *  bins,
int  nDist,
double **  dists,
double *  norm 
) [static]

return the shape LL that a set of values comes from a mixture of tabulated parent distribution.

Makes sure that no bin has a negative expetancy!

Returns:
Log Likelihood
Parameters:
bins list of bin number of input values
nDists number of distributions
dists tabulated distributions
norm normalized weights (i.e. normalized to 1.)
static double Likelihood::shapeLikelihood ( vector< double > *  values,
vector< XeDist * > &  dists,
vector< double > &  weights 
) [static]

return the shape LL that a set of values comes from a mixture of parent distributions.

Returns:
Log Likelihood
Parameters:
values values
dists parent distributions
weights relative weight of distrubtions (needs not be normalized to 1)
static double Likelihood::shapeLikelihood ( vector< double > *  values,
XeDist dist 
) [static]

return the shape LL that a set of values comes from a parent distribution

Returns:
Log Likelihood
Parameters:
values values
dist parent distribution

The documentation for this class was generated from the following file:
 All Classes Functions Variables

Generated on 10 Apr 2015 for Xephyr by  doxygen 1.6.1