rivet is hosted by Hepforge, IPPP Durham
Rivet  2.7.0

Get the e+ e- thrust basis and the thrust, thrust major and thrust minor scalars. More...

#include <Thrust.hh>

Inheritance diagram for Rivet::Thrust:
Rivet::AxesDefinition Rivet::Projection Rivet::ProjectionApplier

Public Member Functions

 Thrust ()
 Constructor.
 
 Thrust (const FinalState &fsp)
 
 DEFAULT_RIVET_PROJ_CLONE (Thrust)
 Clone on the heap.
 
double thrust () const
 
double thrustMajor () const
 The thrust major scalar, $ M $, (thrust along thrust major axis).
 
double thrustMinor () const
 The thrust minor scalar, $ m $, (thrust along thrust minor axis).
 
double oblateness () const
 The oblateness, $ O = M - m $ .
 
const Vector3thrustAxis () const
 
const Vector3thrustMajorAxis () const
 The thrust major axis (axis of max thrust perpendicular to thrust axis).
 
const Vector3thrustMinorAxis () const
 The thrust minor axis (axis perpendicular to thrust and thrust major).
 
const Vector3axis1 () const
 AxesDefinition axis accessors.
 
const Vector3axis2 () const
 The 2nd most significant ("major") axis.
 
const Vector3axis3 () const
 The least significant ("minor") axis.
 
Direct methods

Ways to do the calculation directly, without engaging the caching system

void calc (const FinalState &fs)
 Manually calculate the thrust, without engaging the caching system.
 
void calc (const vector< Particle > &fsparticles)
 Manually calculate the thrust, without engaging the caching system.
 
void calc (const vector< FourMomentum > &fsmomenta)
 Manually calculate the thrust, without engaging the caching system.
 
void calc (const vector< Vector3 > &threeMomenta)
 Manually calculate the thrust, without engaging the caching system.
 
- Public Member Functions inherited from Rivet::AxesDefinition
virtual ~AxesDefinition ()
 Virtual destructor.
 
virtual unique_ptr< Projectionclone () const =0
 Clone on the heap.
 
- Public Member Functions inherited from Rivet::Projection
virtual std::string name () const
 Get the name of the projection.
 
 Projection ()
 The default constructor.
 
virtual ~Projection ()
 The destructor.
 
bool before (const Projection &p) const
 
virtual const std::set< PdgIdPair > beamPairs () const
 
ProjectionaddPdgIdPair (PdgId beam1, PdgId beam2)
 
- Public Member Functions inherited from Rivet::ProjectionApplier
 ProjectionApplier ()
 Constructor.
 
void markAsOwned () const
 Mark this object as owned by a proj-handler.
 
std::set< ConstProjectionPtr > getProjections () const
 Get the contained projections, including recursion.
 
bool hasProjection (const std::string &name) const
 Does this applier have a projection registered under the name name?
 
template<typename PROJ >
const PROJ & getProjection (const std::string &name) const
 
template<typename PROJ >
const PROJ & get (const std::string &name) const
 
const ProjectiongetProjection (const std::string &name) const
 
template<typename PROJ >
const PROJ & applyProjection (const Event &evt, const Projection &proj) const
 Apply the supplied projection on event evt. More...
 
template<typename PROJ >
const PROJ & apply (const Event &evt, const Projection &proj) const
 
template<typename PROJ >
const PROJ & applyProjection (const Event &evt, const PROJ &proj) const
 Apply the supplied projection on event evt. More...
 
template<typename PROJ >
const PROJ & apply (const Event &evt, const PROJ &proj) const
 
template<typename PROJ >
const PROJ & applyProjection (const Event &evt, const std::string &name) const
 
template<typename PROJ >
const PROJ & apply (const Event &evt, const std::string &name) const
 
template<typename PROJ >
const PROJ & apply (const std::string &name, const Event &evt) const
 

Protected Member Functions

void project (const Event &e)
 Perform the projection on the Event.
 
int compare (const Projection &p) const
 Compare projections.
 
- Protected Member Functions inherited from Rivet::Projection
LoggetLog () const
 Get a Log object based on the getName() property of the calling projection object.
 
void setName (const std::string &name)
 Used by derived classes to set their name.
 
Cmp< ProjectionmkNamedPCmp (const Projection &otherparent, const std::string &pname) const
 
Cmp< ProjectionmkPCmp (const Projection &otherparent, const std::string &pname) const
 
virtual Projectionoperator= (const Projection &)
 Block Projection copying.
 
- Protected Member Functions inherited from Rivet::ProjectionApplier
LoggetLog () const
 
ProjectionHandlergetProjHandler () const
 Get a reference to the ProjectionHandler for this thread.
 
template<typename PROJ >
const PROJ & declareProjection (const PROJ &proj, const std::string &name)
 Register a contained projection. More...
 
template<typename PROJ >
const PROJ & declare (const PROJ &proj, const std::string &name)
 Register a contained projection (user-facing version) More...
 
template<typename PROJ >
const PROJ & declare (const std::string &name, const PROJ &proj)
 Register a contained projection (user-facing, arg-reordered version) More...
 
template<typename PROJ >
const PROJ & addProjection (const PROJ &proj, const std::string &name)
 Register a contained projection (user-facing version) More...
 

Detailed Description

Get the e+ e- thrust basis and the thrust, thrust major and thrust minor scalars.

Author
Andy Buckley

The scalar (maximum) thrust is defined as

\[ T = \mathrm{max}_{\vec{n}} \frac{\sum_i \left|\vec{p}_i \cdot \vec{n} \right|}{\sum_i |\vec{p}_i|} \]

, with the direction of the unit vector $ \vec{n} $ which maximises $ T $ being identified as the thrust axis. The unit vector which maximises the thrust scalar in the plane perpendicular to $ \vec{n} $ is the "thrust major" direction, and the vector perpendicular to both the thrust and thrust major directions is the thrust minor. Both the major and minor directions have associated thrust scalars.

Thrust calculations have particularly simple forms for less than 4 particles, and in those cases this projection is computationally minimal. For 4 or more particles, a more general calculation must be carried out, based on the Brandt/Dahmen method from Z. Phys. C1 (1978). While a polynomial improvement on the exponential scaling of the naive method, this algorithm scales asymptotically as $ \mathcal{O}\left( n^3 \right) $. Be aware that the thrust may easily be the most computationally demanding projection in Rivet for large events!

The Rivet implementation of thrust is based heavily on Stefan Gieseke's Herwig++ re-coding of the 'tasso' code from HERWIG.

NB. special case with >= 4 coplanar particles will still fail. NB. Thrust assumes all momenta are in the CoM system: no explicit boost is performed. This can be dealt with by appropriate choice of the supplied FinalState.

Member Function Documentation

◆ thrust()

double Rivet::Thrust::thrust ( ) const
inline

Thrust scalar accessors The thrust scalar, $ T $, (maximum thrust).

◆ thrustAxis()

const Vector3& Rivet::Thrust::thrustAxis ( ) const
inline

Thrust axis accessors The thrust axis.

Referenced by axis1().


The documentation for this class was generated from the following files:
  • include/Rivet/Projections/Thrust.hh
  • src/Projections/Thrust.cc