Page MenuHomeHEPForge
Files F19244230

RhoDMatrixPropagator.h
No OneTemporary
Actions

Size
12 KB
Referenced Files
None
Subscribers
None

RhoDMatrixPropagator.h
View Options

// -*- C++ -*-
#ifndef HERWIG_RhoDMatrixPropagator_H
#define HERWIG_RhoDMatrixPropagator_H
//
// This is the declaration of the <!id>RhoDMatrixPropagator<!!id> class.
//
// CLASSDOC SUBSECTION Description:
//
// This class is reponsible for the computation of the spin density matrix (rho) <BR>
// and decay matrix (D), in terms of the rho and D matrices of the particles <BR>
// (parent, and/or siblings, and/or children) connected to the same "vertex" <BR>
// (hard subprocess, or decay vertex, or splitting vertex) and the amplitude of the <BR>
// "vertex" in terms of the helicities of its connected particles. <BR>
// This class has also a switch that allows to turn OFF the rhoD matrix <BR>
// propagation completely. <BR>
//
// Notice that:
// <UL>
// <LI> there is an unique method <!id>matrixElement(...)<!!id> to calculate <BR>
// the matrix element, and an unique method <!id>computeRhoD(...)<!!id> <BR>
// to update the rhoD matrix, independently from the kind of vertex, <BR>
// whether hard subprocess or decay or splitting. <BR>
// Furthermore, the computation of rhoD matrix is also independent <BR>
// from whether the evolution is forward or backward. <BR>
// These two not evident properties are based on the fact that <BR>
// in <!class>ShowerParticle<!!class> we use a single rhoD matrix to store, <BR>
// in different time, either the spin density matrix rho, or the decay matrix D. <BR>
// This allows a very symmetric and compact writing of the various formulas.
// <LI> the codes of both <!id>matrixElement(...)<!!id> and <!id>computeRhoD(...)<!!id> <BR>
// do not have to separate between the various vertex multiplicity cases: <BR>
// for hard subprocesses: <I> 2 -&GT; 2 , 2 -&GT; 3 , ... , 2 -&GT; N </I> <BR>
// for decays or splittings: <I> 1 -&GT; 2 , 1 -&GT; 3 , ... , 1 -&GT; N </I> <BR>
// The general case of generic <I> N &GT;= 2 </I> is treated directly. <BR>
// This is possible by avoiding to use explicitly nested for loops <BR>
// (which, of course, would require to know exactly how many of them we have <BR>
// to write down) and generating instead automatically all the needed <BR>
// helicity configurations. However, only in the case of splitting processes
// <I> 1-&GT;N </I>, <BR>
// you have to distinguish between the various <I>N</I> in the implementation <BR>
// (but not in the interface) of the method <!id>evaluateAmplitudes(...)<!!id>, <BR>
// using downcasting.
// </UL>
//
// CLASSDOC SUBSECTION See also:
//
// <a href="http:ShowerParticle.html">ShowerParticle.h</a>.
//
#include "Pythia7/Handlers/HandlerBase.h"
#include "ShowerConfig.h"
#include "Herwig++/Config/GlobalParameters.h"
#include "Pythia7/MatrixElement/MEBase.h"
namespace Herwig {
using namespace Pythia7;
class Pythia7::PartialCollisionHandler; // forward declaration
class Pythia7::Decayer; // forward declaration
class RhoDMatrixPropagator: public Pythia7::HandlerBase {
public:
inline RhoDMatrixPropagator();
inline RhoDMatrixPropagator(const RhoDMatrixPropagator &);
virtual ~RhoDMatrixPropagator();
// Standard ctors and dtor.
inline bool isRhoDPropagationON() const;
// This method returns true (false) if the rhoD propagation
// is switched ON (OFF).
double matrixElement( const tMEPtr hardSubME, const CollecShoParPtr & collecShoPar,
const tShoParPtr aParticle );
// Given <!id>aParticle<!!id> coming from a given vertex <BR>
// --- this must be the decaying particle in the case of a decay vertex;
// or the emitting particle in the case of a splitting vertex; or any of the
// particles entering the hard subprocess, in the case of the hard subprocess vertex.
// Notice that only in the latter case the <!id>hardSubME<!!id> is set
// properly (not null) and used to access the hard subprocess matrix element,
// and only in this case the collections of shower particles is needed in order
// to find the ones connected with the hard process --- <BR>
// it computes the real matrix element to be used to generate such vertex.
// If something goes wrong, the method returns 0.
// In the case the <!id>onoffSwitchMode<!!id> is <I>0 (OFF)</I>, the method
// does nothing and returns <I>1.0</I>.
bool computeRhoD( const tMEPtr hardSubME, const CollecShoParPtr & collecShoPar,
const tShoParPtr theParticle );
// It computes the rhoD matrix of <!id>theParticle<!!id> in terms of the rhoD
// matrices of the other particles entering the same vertex, and the
// amplitude (and its conjugate) of the vertex. The "vertex" can be
// a hard subprocess one --- in which case, and only in this one, the
// pointer to the hard subprocess matrix element <!id>hardSubME<!!id>,
// is properly set (not null) and also only in this case the collections
// of shower particles is needed in order to find the ones connected with
// the hard subprocess ---
// or a decay one, or a splitting one.
// The method returns true if it succeed, false otherwise.
// In the case the <!id>onoffSwitchMode<!!id> is <I>0 (OFF)</I>,
// the method does nothing and returns true.
public:
void persistentOutput(PersistentOStream &) const;
void persistentInput(PersistentIStream &, int);
// Standard functions for writing and reading from persistent streams.
static void Init();
// Standard Init function used to initialize the interfaces.
protected:
inline virtual IBPtr clone() const;
inline virtual IBPtr fullclone() const;
// Standard clone methods.
protected:
inline virtual void doupdate() throw(UpdateException);
inline virtual void doinit() throw(InitException);
inline virtual void dofinish();
// Standard Interfaced virtual functions.
inline virtual void rebind(const TranslationMap & trans)
throw(RebindException);
// Change all pointers to Interfaced objects to corresponding clones.
inline virtual IVector getReferences();
// Return pointers to all Interfaced objects refered to by this.
private:
static ClassDescription<RhoDMatrixPropagator> initRhoDMatrixPropagator;
// Describe a concrete class with persistent data.
RhoDMatrixPropagator & operator=(const RhoDMatrixPropagator &);
// Private and non-existent assignment operator.
Complex trace( const ComplexMatrix & matrix ) const;
// Returns the trace of the matrix.
bool normalize( ComplexMatrix & matrix );
// If the trace of the matrix is not zero then it normalizes the matrix
// (that is the new matrix has trace equal <I>1</I>) and returns true;
// otherwise it returns false.
void evaluateAmplitude( const tShoParPtr particle,
const tcPDVector & dataParticles,
const vector<Lorentz5Momentum> & momenta,
const vector<int> & helicities, const vector<int> & helicitiesPrime,
Complex & amplitudeValue, Complex & amplitudePrimeValue );
// We can treat similarly hard <I>2-&GT;N</I> processes
// and decay <I>1-&GT;N</I> processes, because both depend on momenta
// and helicities, but the case of splitting functions must instead be dealt
// differently, because the dependency is, in the common <I>1-&GT;2</I> case,
// on <I>(z,phi)</I> and helicities rather than in momenta and helicities.
// This method receives in input the following parameters:
// --- the pointer to the emitting particle, in the case of
// a splitting process (whereas it is ignored in the other
// cases, hard processes and decay processes);
// --- a vector of data particles for all particles attached to the "vertex"
// (whatever it is: hard process, decay process, splitting process);
// --- momenta of all the particles attached to the "vertex"
// --- the two helicities configurations, helicities and helicitiesPrime;
// and it returns the values of the amplitudes associated to the vertex
// (hard process, decay process, splitting process) in correspondence
// of the two helicities configurations.
// Notice that, whereas for hard <I>2-&GT;N</I> processes and
// decay <I>1-&GT;N</I> processes the procedure holds for any
// final state multiplicity <I>N</I>, in the case of
// splitting <I>1-&GT;N</I> process the interface of this method remains
// unchanged, but in the implementation of this method you have
// to esplicitly distinguish between difference cases, using downcasting
// (see comment in the implementation code).
bool nextIndexConfiguration( const vector<int> & sizes,
vector<int> & indeces, vector<int> & indecesPrime );
// This method determine the next index configuration, given in input
// the vector (sizes) of max values for each index, and the current
// index configuration (vectors: indeces and indecesPrime). The returned
// new configuration is overwritten to the current one.
// The method returns false if there is not anymore other index
// configurations to be considered; true otherwise.
double matrixElement( const CollecShoParPtr & particles );
// Given all of the <!id>particles<!!id> coming from the same vertex, it computes
// the real matrix element to be used to generate such vertex.
// If something goes wrong, the method returns <I>0</I>.
bool computeRhoD( const tShoParPtr theParticle, const CollecShoParPtr & particles );
// Given all of the <!id>particles<!!id> coming from the same vertex, it computes
// the rhoD matrix of one of them, <!id>theParticle<!!id> .
// It returns true if it succeed, false otherwise.
bool setVertexPointer( const tMEPtr hardSubME, const tShoParPtr particlePtr );
// It sets the hard subprocess matrix element pointer, or the decayer pointer,
// or the splitFun pointer depending if the vertex to which the <!id>particlePtr<!!id>
// is connected to is respectively a hard subprocess vertex, a decay vertex,
// a splitting vertex. Notice that in the former case, the input particle,
// <!id>particlePtr<!!id>, can be any of the incoming or outgoing particles
// in the hard subprocess, whereas in the latter two cases the particle must be
// the decaying one or the emitting (splitting) one.
// Notice that in the case of an outgoing particle from the hard
// subprocess, such particle could also be, at the same time, a decaying
// or emitting (splitting) particle, therefore there could be in
// principle an ambiguity about which vertex to consider for it.
// We use the convention that <!id>hardSubME<!!id> is passed not null
// only when we want to consider the hard process as vertex.
// The method returns false if it does not succeed, true otherwise.
void findVertexParticles( const CollecShoParPtr & collecShoPar , const tShoParPtr aParticle,
CollecShoParPtr & particles );
// This method puts in the vector <!id>particles<!!id> all the particles
// (indeed pointers to ShowerParticles objects) that belong to
// the same "vertex" (hard subprocess, or decay, or splitting) as
// the the input particle <!id>aParticle<!!id>. It needs the collection of
// all shower particles only in the case that the vertex is the
// hard subprocess (because the subprocess object would give only
// the incoming and outgoing Pythia7 particles that enter the
// hard subprocess, but not the ShowerParticle objects we want.
// In order to access the amplitude of the "vertex", as a function of
// the helicities of the particles connected to that vertex, we need
// a pointer to either the hard subprocess matrix element, or the decayer,
// or the splitFun. Notice that only one of them should be not null.
tMEPtr _hardSubME;
tDecayerPtr _decayer;
tSplitFunPtr _splitFun;
int _onoffSwitchMode;
};
}
// CLASSDOC OFF
namespace Pythia7 {
// The following template specialization informs Pythia7 about the
// base class of RhoDMatrixPropagator.
template <>
struct BaseClassTrait<Herwig::RhoDMatrixPropagator,1> {
typedef Pythia7::HandlerBase NthBase;
};
// The following template specialization informs Pythia7 about the
// name of this class and the shared object where it is defined.
template <>
struct ClassTraits<Herwig::RhoDMatrixPropagator>: public ClassTraitsBase<Herwig::RhoDMatrixPropagator> {
static string className() { return "/Herwig++/RhoDMatrixPropagator"; }
// Return the class name.
static string library() { return "libHwShower.so"; }
// Return the name of the shared library to be loaded to get
// access to this class and every other class it uses
// (except the base class).
};
}
#include "RhoDMatrixPropagator.icc"
#endif /* HERWIG_RhoDMatrixPropagator_H */

File Metadata

Mime Type
text/x-c++
Expires
Tue, Sep 30, 4:40 AM (22 h, 38 m)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
6527251
Default Alt Text
RhoDMatrixPropagator.h (12 KB)

Event Timeline

HEPForge.org · .