diff --git a/NEWS b/NEWS
--- a/NEWS
+++ b/NEWS
@@ -1,958 +1,963 @@
ThePEG News -*- outline -*-
===============================
Numbered bugs can be found at
https://herwig.hepforge.org/trac/ticket/NN
where NN is the bug number.
The latest version of ThePEG can be found at http://www.thep.lu.se/ThePEG
or at https:///herwig.hepforge.org/
+* ThePEG-2.1.2 release: pending
+
+** Allow LHAPDF6 interface to return photon pdf.
+
+
* ThePEG-2.1.1 release: 2017-07-14
** Added missing evaluate() option to GeneralVVSVertex
** More robust reading of LHE files
** Warn about duplicate PDG names in user input
** Write current timestamp into log files
** Calling 'read' in an input file will no longer change the
repository directory you're in
* ThePEG-2.1.0 release: 2017-05-19
** Transition to C++-11 code, building with C++-98 is no longer possible
** Interfaces with two choices now consistently accept Yes/No as answer
** Several smaller bug fixes and additions, to allow new features in Herwig 7.1
* ThePEG-2.0.4 release: 2016-10-25
** Default weight explicitly labelled
The nominal event weight is now labelled "Default" and is always the
first weight to be inserted into HepMC.
(Note that HepMC 2.06.09 may not preserve this ordering.)
** LesHouches event weights handling
There are now two options for the treatment of optional weights
coming in from a Les Houches file, available through
"LesHouchesEventHandler:WeightNormalization". Either they are
normalized to the "Default" weight ("Normalized"), or they are given
as cross-sections in pb ("CrossSection"). The default behaviour
is "Normalized".
** JetFinder in Cuts objects writes information
The selected JetFinder is explicitly listed in the debug output
* ThePEG-2.0.3 release: 2016-07-21
** LesHouches reader
Removed FxFx-specific weights from default output
** FuzzyTheta cut
Fixed missing division by width
** NaN check for incoming momenta from ME providers
** Build system
Compatibility with gcc-6, improved configure macros.
* ThePEG-2.0.2 release: 2016-04-27
** LesHouches reader
Fixed cross-section issue for LH event files using weight scheme 3
* ThePEG-2.0.1 release: 2016-02-17
** Electroweak scheme fixes
Schemes 2 and 5 now set GF, scheme 7 calculates mw correctly.
** New ParVector 'clear' command
** More visible version information
The version number of ThePEG is written into the log file.
** Rivet exception handling
Any exceptions from Rivet/YODA are converted into warning exceptions.
This can invalidate plots, please check the logfile.
** Rapidity limits fixed in FuzzyTheta cuts
** Doxygen updated to version 1.8
** Java interface fixed
* ThePEG-2.0.0 release: 2015-12-04
** Improvements for NLO
A number of improvements and bug fixes have been made in the course of
(Herwig 7) NLO development.
** Improvements for spin correlations
A number of improvements have been made to handle spin correlations in the
parton shower.
** Setup file mechanism
An isolated event generator can now be modified prior to generating events
by providing an additional input file; also support for re-initalizing
isolated event generators has been added in this course to support Herwig
7's parallel integration modes.
** Handling of weighted LHE events
The handling of LHE events with variable weight (codes 3 and 4) has now
been enabled, along with the full handling of multiple event weights down
to the HepMC output.
** LHAPDF6 handling
A problem in initializing LHAPDF6 has been fixed.
** Heavy ion collisions
A dedicated HepMC interface for heavy ion collisions has been introduced.
** First steps for unit testing
Unit testing has been introduced for a number of components based on
boost's unittesting framework. They are only enabled, when configured
--with-boost .
* ThePEG-1.9.2 release: 2014-07-07
** Better support for LHAPDF 6
The LHAPDF interface now determines if version 6 is available and
makes use of new LHAPDFv6 features.
* ThePEG-1.9.1 release: 2014-04-30
** Build fix for SLC6
Fixed problems with a missing <algorithm> header in ACDCGen.h that affected
some builds on SLC6
** Build fix for Rivet 2.1.1
Adapted to Rivet 2.1.1 changed header file layout
** OS X Mavericks build fix for readline
The problem of infinite hangs when linking against libreadline
in Mavericks has been resolved.
* ThePEG-1.9.0 release: 2013-10-28
** Rapidity edge cases
Changed behavior for vanishing pt in eta() and vanishing mt in
rapidity(): no error is thrown, but a ridculously large rapidity is
returned instead.
** Command line interface
Adding a tag of the form "#first-last" when running a
MultiEventGenerator now allows to run a subset of the runs
** Pre- and Posthandlers
A warning is now issued if the same handler is inserted twice into
a pre- or posthandler list.
** Statistics fixes
Various issues for the handling of statistics in the presence of
negative weights have been fixed.
** Rivet search paths
A new interface RivetAnalysis:Paths allows more search paths to be
added to the analysis finding routine. This supplements the
RIVET_ANALYSIS_PATH environment variable.
** Vertex classes
A chain of sign and prefactor fixes has been implemented in the
Vertex classes to correct various vertex structures in Herwig++
conversions of Feynrules models.
** Custom XComb objects
Matrix element classes can built with customized XComb objects.
** Reweighting of SubProcessGroups and projections
Matrix element groups can reweight contributions in a
SubProcessGroup and select one of the dependent subprocesses to be
used as the hard one instead of the group in total.
** Fuzzy cuts
Jet cuts in the jet cut framework can be assigned a fuzzy shape to
improve the stability of NLO calculations.
** NLORivetAnalysis, NLOHepMCFile
The NLOHepMCFile and NLORivetAnalysis classes have been added to
allow Rivet to analyse NLO calculations which perform output for the
subtracted real emission matrix elements as ThePEG::SubProcessGroups;
individual subprocesses in these groups are flagged as correlated
by being assigned the same event number.
** C++-11 testing
To help with the coming transition to C++-11, we provide the new
--enable-stdcxx11 configure flag. Please try to test builds with
this flag enabled and let us know any problems, but do not use this
in production code yet. In future releases, this flag will be on by
default.
* ThePEG-1.8.3 release: 2013-02-22
** Fixed HepMC status code assignment
The incoming line to a technical vertex that only changes a particle's
momentum is now labelled 11 (MC-internal) instead of 2 (unstable).
* ThePEG-1.8.2 release: 2013-01-30
** Changes to scheduled event dumps
Set the EventGenerator option "KeepAllDumps" to keep all dump files
of a run, labelled by event number. These scheduled dumps now produce
a cleaned generator state between events.
** SLHA block bug fix
In some circumstances, the end of an SLHA block was not parsed
correctly.
* ThePEG-1.8.1 release: 2012-10-15
** Repository changes
*** Stable flag
The behaviour of the 'Particle:Stable' flag has changed slightly. When
all decaymodes are removed from a particle, the flag will be
automatically set to stable. However, it still needs to be set unstable
by hand when new decaymodes are added to a formerly stable particle.
*** Search paths
The search paths for reading input files are now saved in the
repository.
** Rivet analysis YODA capable
The configure time check for Rivet will now detect the version to
check if it is YODA capable.
** NLOHepMCFile
An AnalysisHandler, similar to the standard HepMCFile has been
added to allow analysing plain NLO calculations by communicating
each member of a subprocess group (typically real emission and
subtraction contributions) to a HepMC file; until Rivet supports
the correct treatment of errors in this case, the subprocess group
members are written out as events independent of each other.
** Jet cuts
A more flexible framework for cuts on jets has been added,
including jet finders (either builtin or via an optional link to
fastjet). All kinds of exclusive and inclusive jet cuts, as well as
jet vetoes on the level of the hard process are now supported.
** Tree2toNDiagram
A bug has been fixed in Tree2toNDiagram::isSame(tcDiagPtr,
map<int,int>&) method to take into account symmetries when swapping
two external legs attached to the same vertex
** Quark thresholds in alphaS
Support for setting quark masses for threshold matching in
couplings deriving from AlphaSBase independently of masses used in
ParticleData objects has been added. The O1AlphaS implementation is
fully aware of this enhancement.
** LesHouches reader fixes
The LesHouches readers have been improved, and will spot many more
consistency errors, such as coloured Standard Model leptons, or helicity
values outside of [-1,1] and 9.
** Spinor enhancements
The LorentzSpinor class has gained member functions for projection operations
and the contraction with the sigma-mu-nu commutator.
* ThePEG-1.8.0 release: 2012-05-21
** NLO support
*** ThePEG now includes structures to ease implementing next-to-leading
order (NLO) calculations as well as for interfacing external
matrix element codes through runtime interfaces. Particularly,
the newly introduced MEGroup and accompanying StdXCombGroup,
StdDependentXComb and SubProcessGroup classes provide the
functionality required by subtraction approaches to higher
orders. A general interface for cutting on reconstructed jets
as required by higher-order calculations is included, along
with an implementation of kt, Cambridge-Aachen and anti-kt jet
finding as relevant for NLO calculations. Hard process
implementations deriving from MEBase are no longer limited to
the evaluation of PDFs by PartonExtractor objects, thus
allowing for a more flexible and more stable implementation of
finite collinear contributioins appearing in the context of
higher order corrections.
*** The generation of phasespace points
for the hard subprocess has been made more flexible,
particularly to allow generation of incoming parton momenta by
the hard matrix element as is typically done by phasespace
generators provided with fixed-order codes. Along with this
change, generation of the phasespace point does not need to
take place in the centre-of-mass system of the incoming
partons.
*** Various helpers have been added to MEBase and dependent
code along with the improvements described above, including
simple functionality required for caching intermediate results.
*** Tree2toNDiagram supports merging of two external legs,
yielding another Tree2toNDiagram object to assist in
determining subtraction terms required for a particular process
** Support for SU(3)-sextet colour lines
** Named weights support, also in HepMC
Named, optional weights on top of the usual event weight are now
fully supported; this includes their communication to HepMC events
as well as their parsing from the extented Les Houches file format
drafted at the Les Houches workshop 2009.
** Complex masses supported in Helicity code
** Several minor fixes and enhancements
* ThePEG-1.7.3 release: 2012-03-05
The only changes are in LesHouches.so, now at version 14.
** Spin information
Spin correlation information will now be set up correctly for tau
leptons. To go back to the old behaviour, set
LesHouchesReader:IncludeSpin No
** Consistency checks
Catch broken input where mother-daughter relations are circular.
* ThePEG-1.7.2 release: 2011-11-01
** HepMC configuration
Clarified at configure time that HepMC versions before 2.05 are not
officially supported.
** Rivet configuration
Rivet builds with external header dependencies are now correctly
recognized.
** Helicity vertex consistency
To help debugging, the addToList() function for registering
particle lines connected to a vertex now checks for electric charge
conservation at the vertex. Additionally, the specified QED/QCD
order of the interaction is checked.
** Ticket #355: Global library list for resume functionality
The list of global libraries is now correctly included in the dump
file, to fix the functionality of resuming a run from regular
checkpoints.
* ThePEG-1.7.1 release: 2011-06-20
** Ticket #238: Self-consistent electroweak schemes
The default behaviour of Herwig++ is to use the best values for all
the electroweak parameters, which can be inconsistent. Optionally
now, a self-consistent electroweak scheme can be chosen in the
parameter Model:EW/Scheme. Note that values of 'EW/Scheme' away
from the default have not received the same amount of testing.
** Ticket #338: Fixed reporting of floating point exceptions
This was an issue whenever LHAPDF libraries were linked in.
** Fixed cTau() behaviour
The new behavior is that if both width and lifetime are zero,
cTau() returns zero for unstable particles and MaxLength for
stable ones.
** MaxErrors
The cutoff can be disabled by setting MaxErrors to -1.
** StdOut redirect
CurrentGenerator::Redirect now does not redirect to to the internal
stream in EventGenerator if the useStdout flag has been set.
** Run name tags
Aded possibility to add a tag to the run name when running with the
'-t' option. One run file can thus be run with different seeds and
result in different output files.
** Exception names
EventGenerator tries to convert exception type names into
human-readable form. Currently this only works for gcc-compatible
compilers.
** Repository API changes
Instead of printing an error message to cerr, the
Repository::load() and Repository::read(filename,os) commands now
behave like the other repo commands and return an error string.
This allows --exitonerror to work correctly for load() and read().
Users of these functions need to send the string to cerr themselves
if the old output behaviour is required.
Repository::read(is, os, prompt) is unchanged.
** HepMC precision
The precision() option for HepMC GenEvent is now available as an
interface in the HepMCFile analysis handler.
** gcc-4.6
The build has now also been tested with gcc 4.6.0.
* ThePEG-1.7.0 release: 2011-02-08
** Behaviour
*** Cross-section information
The .out file contains a better estimate of the cross-section
directly from the phase space sampler. It should be reliable if
not too many events were vetoed during the later phases of the
production.
*** Simpler decay mode selection
Instead of having to turn off every mode individually, the new
'SelectDecayModes' interface allows commands such as
tbar:SelectDecayModes none
tbar:SelectDecayModes tbar->nu_ebar,e-,bbar; tbar->nu_mubar,mu-,bbar;
Use 'PrintDecayModes' to list the available choices.
*** Rivet interface
The interface will check before the run if all chosen analyses are
actually available in Rivet, and will only call finalize() if any
events have been generated. Event weights are now passed correctly
into Rivet.
*** Les Houches QNUMBER support
QNUMBER particle creation support has been added to the Les
Houches reader.
*** Debug level
If a debug level is set on the command line, it will always be used.
*** Abort preserves events
A hard abort exception during event generation will try to finalize as
best as it can, thus preserving information about the run until this
point.
*** Progress log
There is a new ThePEG::ProgressLog analysis handler, which prints
timing information about the run to screen.
*** Graphviz
The event graph can show missing momentum information.
*** CRLF line endings (Windows-style)
File readers can now cope with files that have CRLF (Windows) and CR
(Mac) line endings.
** Structure
*** Diffraction support
To provide support for the simulation of diffractive events, the
LeptonLeptonRemnant class was renamed to UnresolvedRemnant, the
WeizsackerWilliams PDF extended, BudnevPDF added, and PartonExtractor
modified to allow separate override PDFs for each beam.
*** Polarized beams
The PolarizedBeamParticleData class permits a spin polarization choice
on the incoming particles.
*** Mixing particles
The MixedParticleData class supports mixed particle states.
*** SpinBase
The SpinBase class has been removed, SpinInfo is now the base class.
*** Vertex classes
Several new vertices were added for BSM physics, and some minor bugs
fixed with existing classes.
*** PID type
Particle IDs now have their own type, 'PID'. It can only be converted
to 'long' to avoid unsigned int overflow errors on 64bit machines.
*** ClassDescription
There is a simpler way to register ClassDescription information with
the Repository: the DescribeClass<> template. No information is needed
in the class headers anymore, reducing build dependencies. New code
will use this method now, older classes will be migrated in future.
*** LWH histogramming
The built-in implementation of AIDA histogramming has been
restructured slightly to decouple the implementation from the
interface.
** Build system
*** Silent build rules
'make' now builds silently, to improve readability. To get the old
behaviour, run 'make V=1'
*** zlib
Integrated zlib to read compressed files, such as Les Houches events.
*** Libtool 2.4
Will fix some issues on OS X.
*** gsl check
This check now also works with Rivet
** Bug fixes
*** Ticket #286: Implementation of Cuts in LesHouchesReader
Les Houches events are given in the lab frame, but QCDCuts expects to
be given momenta in the parton-parton cmf. Boost added.
*** Ticket #303: AlphaS thresholds
AlphaS thresholds were fixed for exactly massless quarks.
*** Ticket #304: Length units in HepMC
*** Ticket #309: Particle initialization order
*** Ticket #310: LeptonLeptonPDF fix
* ThePEG-1.6.1 release: 2009-12-11
** Ticket #292: Spin correlation fix to stabilize tau decay numerics
We have restructured the spin correlation code to stabilize the
numerical problems seen in tau decays.
** Speed increase
Improved decay chain handling speeds up a typical run by a few percent.
** Exception logging
The log file lists different exception types individually again
instead of grouping them all by severity only. This was broken in
the last few releases.
* ThePEG-1.6.0 release: 2009-11-19
** Helicity amplitudes
The main change in this release is a streamlining of the helicity
amplitude code. If you have self-written vertex classes, you will
need to adapt them slightly:
*** Spinor representation
All spinors are now in HELAS representation, the optional
representation switching has been removed.
*** Adding particles to the vertex
Instead of calling setList(v1, v2, v3) with three vectors of
particles that need to be filled in sync, call
addToList(p1, p2, p3) repeatedly.
*** Name changes
To make them consistent with the rest of ThePEG, the names of
functions starting with get...() have changed:
getFoo() becomes foo(); setFoo(...) becomes foo(...).
** Vector code changes
*** Vector3 has been renamed ThreeVector
This makes it consistent with the other vector classes.
*** LorentzVector::mag()/mag2() removed, to reduce confusion
Use the equivalent ::m()/m2() instead to get E^2-p^2.
LorentzVector::rho2() == ThreeVector::mag2() == p^2
*** Calculating with a zero-length vector
Previously, mathematically undefined values were arbitrarily set
to 0 in this situation. Now an assertion failure is triggered in
debug mode. Only the azimuthal angle phi() still returns 0.
** Environment variable interpretation removed
The programs setupThePEG and runThePEG are not wrapped in shell
scripts anymore. All usage of environment variables at runtime has
been removed. To influence the behaviour of ThePEG, you will need
to use explicit command line flags, or calls to Repository::
functions.
** StandardModelBase now provides G_Fermi
InvEnergy2 StandardModelBase::fermiConstant() const;
provides the PDG 2006 value of G_Fermi.
** Allow particle width cuts to be unset
Setting a negative value for the lower or upper width cut removes
that bound on the width.
** Ticket #271: Decay mode names are normalized
Previously, the ordering of decay products in a decay mode
specifier needed to match ThePEG's internal ordering exactly. This
is now done automatically.
** Ticket #273: NoPDF and LesHouches reader
The LesHouches reader has been fixed for the case where no PDFs are
used in the LHE file.
** Ticket #275: doinitrun() ordering
The ACDC sampler now ensures that initrun() of all ME objects is
run first.
** Ticket #277: Repository command help texts
ThePEG's repository command language now includes a 'help'
functionality.
** Redirection of various files
The log file stream now goes to stdout instead of stderr when
EventGenerator:UseStdout is set.
** Readline support can be disabled
Using the configure switch '--disable-readline', the linking of
libreadline can be suppressed. This can be useful for batch code that
will never be used interactively, if the number of linked-in
libraries is a problem.
** --with-LHAPDF configure flag
Previously, the full LHAPDF path including lib/ needed to be
specified. This now also works with the more common usage of just
LHAPDF's prefix path.
** Rivet analysis plugin
The Rivet analysis output files are now named consistently like the
other output from a run, with the run name as prefix.
** Graphviz event visualization
This is now independent of HepMC. The call to
void ThePEG::printGraphviz(ostream & os, tcEventPtr event);
on any event will output a Graphviz file to the stream,
suitable for interpretation with the 'dot' tool. It shows a
visualization of the generator's internal event structure, useful
for debugging. This is an initial version, feedback is welcome!
** Fixed compatibility with older HepMC versions
A problem with rejecting a missing HepMC unit implementation was fixed.
* ThePEG-1.5.0 release: 2009-09-01
** New ZERO object
The ZERO object can be used to set any dimensionful quantity to
zero. This avoids explicit constructs like 0.0*GeV.
** Readline interface
The interactive repository access now uses the readline library,
providing a command history and easier command line editing.
** Syntax highlighting
We now have a syntax highlighting mode for emacs. To enable it, use
'M-x ThePEG-repository-mode' on any .in file.
** Configure information
Important configuration information is listed at the end of the
'configure' run and in the file 'config.thepeg'. Please provide
this file in any bug reports.
** Rivet interface
ThePEG now supports Rivet internally as an AnalysisHandler if it's
available.
** HepMC tools; CLHEP support removed
The HepMC file output and graphviz event view have migrated from
Herwig++ to ThePEG. The deprecated CLHEP support has been removed.
** Exception specifiers removed
Client code changes are needed in doinit() etc. Simply remove the
exception specifier after the function name.
** Support for HepMC 2.05
*** New features
ThePEG now supports cross-section output, PDF information and unit
specifications if they are available in HepMC.
*** IO_ASCII
Support for the deprecated IO_ASCII format has been removed.
*** Status codes
ThePEG uses codes 1, 2 and 11 according to the HepMC agreement.
** Redirection of .out, .log and .tex to stdout
Set 'EventGenerator:UseStdout' to 'Yes' and (almost) all output
will be streamed to stdout instead of files.
** Ticket #248: Les Houches reader
The cross-section information is now reported correctly when
reading several files.
** Cuts output
If the debug level is set > 0, the current set of cuts is prepended
to the logfile.
** Preweighting of matrix elements
A segfault when using preweights was fixed. Preweights are now
correctly included in handler statistics.
** Other technical changes
*** Colour line improvements
*** PDFsets.index search improved
*** Ticket #232: Java check on OS X now works headless
*** Running couplings restructured
*** LeptonLeptonRemnant iprovements to support GammaGamma
*** WaveFunction constructors streamlined
*** VertexBase now provides sin2ThetaW etc.
* ThePEG-1.4.2 release: 2009-05-08
** Ticket #242
Fixed a compiler problem that showed up on openSUSE 10.2, g++
4.1.2. A source line was omitted if the optimization level was
higher than -O1.
** User interaction
Dump file generation can now be disabled completely by setting
EventGenerator:DumpPeriod to -1.
* ThePEG-1.4.1 release: 2009-03-31
** User interaction
Error messages have been clarified in BaseRepository and
StandardEventHandler
** Les Houches files
File readers are more robust, with clearer messages when things go
wrong.
** Floating point issues
fixed in ThreeVector and VertexBase.
** HepMC converter
Fixed PDF choice for asymmetric beams.
** Libtool
Updated to version 2.2.6
* ThePEG-1.4.0 release: 2008-12-02
** Efficiency improvements
The LorentzVector class, the helicity amplitude code and PDF
lookups have been profiled and restructured to eliminate speed
bottlenecks.
** Deep inelastic scattering
Support for DIS is now implemented in ThePEG.
** New rapidity cut
Added alternative pT cut that cuts on rapidity rather than
pseudorapidty as the existing one fails for zero-pt massive
particles.
** HepMC units support
Users of HepMC versions > 2.04 get full units support. The HepMC
GenEvent object has the correct units set.
** Support for HepMC PdfInfo data
Users of the HepMC converter can now fill the PdfInfo data, which
has been available since HepMC 1.28.00. Older versions are no
longer supported.
** Ticket #199: Particle lifetime cutoff
Users can set a maximum lifetime. Particles that live longer than
this are not decayed.
** Ticket #215: Madgraph
Fixed a problem in reading certain Madgraph event files.
** Les Houches files
Optional rescaling of energy or mass of particles which were read
in. Check the file to trap 'nan' and 'inf' values early during the read-in.
** File cleanup
Most inline functions are now defined in the headers rather than a
separate .icc file.
* ThePEG-1.3.0 release: 2008-06-20
** Statistical errors
Error estimates on the cross-sections are now reported in the .out files.
** Decaymode setup
The decaymode setup has been reworked, keeping backwards
compatibility. The 'defaultparticle' command has been removed.
** Madgraph reader
Updated to latest Madevent version.
** LHAPDF improvements
lhapdf-config is not used anymore to determine the location of the
PDF sets. Instead, they are fixed at configure-time.
** HepMC
The beam particles are now set correctly.
** Arbitrary search paths for .in files
Input files don't have to be in '.', use '-I' to specify additional
directories.
** Ticket #172
Fix for cuts in mirrored processes.
** Simpler emacs macros
The emacs macros have been significantly cleaned up.
** Helicity vertex classes
Potential prolems with uninitialized variables were fixed. No
actual bugs had occurred from here.
** Memory leak fixes and performance
Several loops of shared pointers were fixed involving DecayModes
and ParticleData. FixedSizeAllocator was removed, the regular 'new'
and 'delete' is now used for all allocations, giving a 5% speedup.
* ThePEG-1.2.0 release: 2008-04-18
** ThePEG uses GSL
The GNU Scientific Library and its headers are now required for
building ThePEG. RandomGenerator partially uses GSL now.
** Ticket #160: HepMC converter
Optionally, the HepMC converter can fill an external GenEvent object,
instead of newing one internally.
** 'globallibrary' command
The 'globallibrary' command can be used to register libraries which
are useful for the whole run. They do not need to be listed in each
class's library() function.
** Resume from dump files (also solves #149)
The --resume command line flag instructs runThePEG to resume the
run from a previous dump file. 'set Generator:DumpPeriod 1000'
writes such a checkpoint every 1000 events.
** New Repository interface (also solves #141)
The repository now provides a Repository::cleanup() method, to be
called at the end of a run. Alternatively, a Repository object can
be created and its member functions called.
** Les Houches interface improvements.
LesHouchesReader had had a major overhaul.
** XSecCheck analysis
Issues a warning if a target cross section is not met in the run.
** Weighted events
Handling of weighted events was improved.
** Ticket #124: 'Deleted' interface option
Interfaces can be declared obsolete, a warning will be issued when
they're used.
** LHAPDF interface
The interface now allows photons as partons inside a hadron.
** gcc 4.3.0
ThePEG compiles cleanly with the new gcc 4.3 series.
** Bug #138
Lepton NoPDF now works.
* ThePEG-1.1.2 release: 2008-02-25
** Bug #136: dSigHatDR
Efficiency improvement for zero PDF.
** Bug #137: HepMC conversion
The HepMC converter now takes a unit argument to specify which
energy and length units should be used in the HepMC event.
** Bug #140 / #141: Crash on shutdown
This fixes a bug introduced in 1.1.1. External code interfaces
should now work again.
** Bug #151: Loop in Remnant record
The loop between remnants in the event record has been removed.
** PDF improvements
Fix to handling of maximum flavour from LHAPDF. Fix for xmin
calculation.
* ThePEG-1.1.1 release: 2007-12-07
** Bug #46: Reproducibility
Fixed a problem where runs were not identical for a given random
number seed. You now _must_ reset the seed if you need independent
event generator runs.
** Detection of gcc abs bug
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34130
configure now checks for and works around the gcc abs() bug.
** Separate LWHFactory library
Fixed problem in Rivet interaction by factoring out LWHFactory into
its own dynamic library.
* ThePEG-1.1.0 release: 2007-11-20
** New vector classes
ThePEG now uses its own internal Vector classes, making it
independent of CLHEP.
** New dimension checking mechanism
Optionally, any physical expression is checked for dimensional
correctness at compile time.
** Extended Helicity classes
A full set of helicity amplitude classes has been transferred from
Herwig++.
** unlisted
Many other improvements and small bug fixes, see ChangeLog.
* ThePEG-1.0.1 release: 2006-11-22
** unlisted:
Fixed memory leak in LesHouchesReader.
** Bug #58
maximumCMEnergy() member of the EventGenerator returns zero. See
ChangeLog entry 2006-10-06
** Bug #62
fixed 'const' behaviour in Lorentz spinor classes
** Bug #68
Improved error message for switch options
** unlisted
Improved compile-time LHAPDF library and include file handling.
** unlisted
Bug in IteratorRange::rrange(const Container &).
** unlisted
fixed Selector::swap()
** unlisted
Bug in ClusterCollapser where no colour-singlet particles were
considered for momentum compensation if no coloured particles were
present.
** unlisted
Bug in LeptonLeptonRemnant: minX variable not persistent
** unlisted
scale of the produced coloured particles was not set in
Onium3GDecayer and ColourPairDecayer
** unlisted
unused default path removed from DynamicLoader
* ThePEG-1.0 release: 2006-09-27
diff --git a/PDF/LHAPDF6.cc b/PDF/LHAPDF6.cc
--- a/PDF/LHAPDF6.cc
+++ b/PDF/LHAPDF6.cc
@@ -1,478 +1,470 @@
// -*- C++ -*-
//
// LHAPDF6.cc is a part of ThePEG - Toolkit for HEP Event Generation
// Copyright (C) 2014 Leif Lonnblad, David Grellscheid
//
// ThePEG is licenced under version 3 of the GPL, see COPYING for details.
// Please respect the MCnet academic guidelines, see GUIDELINES for details.
//
//
// This is the implementation of the non-inlined, non-templated member
// functions of the LHAPDF class.
//
#include "LHAPDF6.h"
#include "ThePEG/Interface/ClassDocumentation.h"
#include "ThePEG/Utilities/DescribeClass.h"
#include "ThePEG/Interface/Parameter.h"
#include "ThePEG/Interface/Switch.h"
#include "ThePEG/Interface/Command.h"
#include "ThePEG/Interface/Deleted.h"
#include "ThePEG/PDT/EnumParticles.h"
#include "ThePEG/PDT/ParticleData.h"
#include "ThePEG/Utilities/Throw.h"
#include "ThePEG/Persistency/PersistentOStream.h"
#include "ThePEG/Persistency/PersistentIStream.h"
#include "LHAPDF/LHAPDF.h"
using std::vector;
using std::string;
using std::pair;
using ThePEG::GeV2;
using ThePEG::cPDVector;
ThePEG::LHAPDF::LHAPDF()
: thePDF(), thePDFName("THEPEG_NO_PDFSET_CHOSEN"),
theMember(0), theMaxFlav(5),
xMin(0.0), xMax(1.0), Q2Min(ZERO), Q2Max(Constants::MaxEnergy2) {}
ThePEG::LHAPDF::LHAPDF(const LHAPDF & x)
: PDFBase(x),
thePDF(), thePDFName(x.thePDFName),
theMember(x.theMember),
theMaxFlav(x.theMaxFlav),
xMin(x.xMin), xMax(x.xMax), Q2Min(x.Q2Min), Q2Max(x.Q2Max) {}
ThePEG::IBPtr ThePEG::LHAPDF::clone() const {
return new_ptr(*this);
}
ThePEG::IBPtr ThePEG::LHAPDF::fullclone() const {
return new_ptr(*this);
}
void ThePEG::LHAPDF::initPDFptr() {
::LHAPDF::setVerbosity(std::max(0,
ThePEG::Debug::level - 1));
if ( thePDF
&& thePDF->set().name() == thePDFName
&& thePDF->memberID() == theMember )
return;
delete thePDF;
thePDF = ::LHAPDF::mkPDF(thePDFName, theMember);
xMin = thePDF->xMin();
xMax = thePDF->xMax();
Q2Min = thePDF->q2Min() * GeV2;
Q2Max = thePDF->q2Max() * GeV2;
}
void ThePEG::LHAPDF::doinit() {
PDFBase::doinit();
initPDFptr();
}
void ThePEG::LHAPDF::dofinish() {
PDFBase::dofinish();
delete thePDF;
thePDF = 0;
}
void ThePEG::LHAPDF::doinitrun() {
PDFBase::doinitrun();
initPDFptr();
}
void ThePEG::LHAPDF::setPDFName(string name) {
if ( ::LHAPDF::endswith(name, ".LHgrid") ) {
name = name.substr(0, name.size() - 7);
}
else if ( ::LHAPDF::endswith(name, ".LHpdf") ) {
name = name.substr(0, name.size() - 6);
}
// fix the eternal typo
if ( name == "cteq6ll" ) name = "cteq6l1";
if ( ::LHAPDF::contains(::LHAPDF::availablePDFSets(), name) ) {
thePDFName = name;
theMember = 0;
}
else {
Throw<ThePEG::LHAPDF::NotInstalled>()
<< "'set " << fullName() << ":PDFName "
<< name << "': PDF not installed. Try 'lhapdf install'.\n"
<< Exception::setuperror;
}
}
void ThePEG::LHAPDF::setPDFMember(int member) {
try {
::LHAPDF::PDFInfo * test =
::LHAPDF::mkPDFInfo(thePDFName, member);
if ( test )
theMember = member;
delete test;
}
catch (::LHAPDF::ReadError & e) {
Throw<ThePEG::LHAPDF::NotInstalled>()
<< e.what() << Exception::setuperror;
}
}
string ThePEG::LHAPDF::doTest(string input) {
double x = 0;
Energy2 Q2 = ZERO;
Energy2 P2 = ZERO;
istringstream is(input);
is >> x >> iunit(Q2, GeV2) >> iunit(P2, GeV2);
initPDFptr();
ostringstream os;
for ( int i = 0; i < 13; ++i ) os << " " << thePDF->xfxQ2(i,x,Q2/GeV2);
return os.str();
}
bool ThePEG::LHAPDF::canHandleParticle(tcPDPtr particle) const {
using namespace ParticleID;
return abs(particle->id()) == pplus || abs(particle->id()) == n0;
}
cPDVector ThePEG::LHAPDF::partons(tcPDPtr particle) const {
// We assume that all standard partons can be extracted.
const ::LHAPDF::PDFSet & pdfset = ::LHAPDF::getPDFSet(thePDFName);
const vector<int> & flavs =
pdfset.get_entry_as< vector<int> >("Flavors");
cPDVector ret;
ret.reserve( flavs.size() );
if ( canHandleParticle(particle) ) {
for ( size_t i=0; i < flavs.size(); ++i ) {
ret.push_back(getParticleData(flavs[i]));
}
}
assert( ret.size() == flavs.size() );
return ret;
}
double ThePEG::LHAPDF::xfx(tcPDPtr particle, tcPDPtr parton,
Energy2 partonScale,
double x, double, Energy2) const {
// Here we should return the actual density.
using namespace ThePEG::ParticleID;
double Q2 = partonScale/GeV2;
if ( ! thePDF->inRangeXQ2(x, Q2) ) {
switch ( rangeException ) {
case rangeThrow: Throw<Exception>()
<< "Momentum fraction (x=" << x << ") or scale (Q2=" << Q2
<< " GeV^2) was outside of limits in PDF " << name() << "."
<< Exception::eventerror;
break;
case rangeZero:
return 0.0;
case rangeFreeze:
x = min(max(x, xMin), xMax);
Q2 = min(max(Q2, Q2Min/GeV2), Q2Max/GeV2);
}
}
int pid = parton->id();
int abspid = abs(pid);
switch ( pid ) {
case t:
case tbar:
case b:
case bbar:
case c:
case cbar:
return maxFlav() < abspid ? 0.0 : thePDF->xfxQ2(pid,x,Q2);
case s:
case sbar:
return thePDF->xfxQ2(pid,x,Q2);
case u:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(d ,x,Q2);
case pbarminus: return thePDF->xfxQ2(ubar,x,Q2);
case nbar0: return thePDF->xfxQ2(dbar,x,Q2);
case pplus:
default: return thePDF->xfxQ2(u ,x,Q2);
}
case ubar:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(dbar,x,Q2);
case pbarminus: return thePDF->xfxQ2(u ,x,Q2);
case nbar0: return thePDF->xfxQ2(d ,x,Q2);
case pplus:
default: return thePDF->xfxQ2(ubar,x,Q2);
}
case d:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(u ,x,Q2);
case pbarminus: return thePDF->xfxQ2(dbar,x,Q2);
case nbar0: return thePDF->xfxQ2(ubar,x,Q2);
case pplus:
default: return thePDF->xfxQ2(d ,x,Q2);
}
case dbar:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(ubar,x,Q2);
case pbarminus: return thePDF->xfxQ2(d ,x,Q2);
case nbar0: return thePDF->xfxQ2(u ,x,Q2);
case pplus:
default: return thePDF->xfxQ2(dbar,x,Q2);
}
case g:
return thePDF->xfxQ2(g,x,Q2);
case ParticleID::gamma:
- return 0.;
+ return thePDF->xfxQ2(ParticleID::gamma,x,Q2);
}
return 0.0;
}
double ThePEG::LHAPDF::xfvl(tcPDPtr particle, tcPDPtr parton,
Energy2 partonScale,
double l, Energy2 particleScale) const {
using Math::exp1m;
return xfvx(particle, parton, partonScale,
exp(-l), exp1m(-l), particleScale);
}
double ThePEG::LHAPDF::xfvx(tcPDPtr particle, tcPDPtr parton,
Energy2 partonScale,
double x, double, Energy2) const {
// Here we should return the actual valence density. This will only
// work properly for nucleons
using namespace ThePEG::ParticleID;
double Q2 = partonScale / GeV2;
if ( ! thePDF->inRangeXQ2(x, Q2) ) {
switch ( rangeException ) {
case rangeThrow: Throw<Exception>()
<< "Momentum fraction (x=" << x << ") or scale (Q2=" << Q2
<< " GeV^2) was outside of limits in PDF " << name() << "."
<< Exception::eventerror;
break;
case rangeZero:
return 0.0;
case rangeFreeze:
x = min(max(x, xMin), xMax);
Q2 = min(max(Q2, Q2Min/GeV2), Q2Max/GeV2);
}
}
switch ( parton->id() ) {
case t:
case tbar:
case b:
case bbar:
case c:
case cbar:
case s:
case sbar:
case ParticleID::gamma:
return 0.0;
case u:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(d,x,Q2) - thePDF->xfxQ2(dbar,x,Q2);
case pbarminus: return 0.0;
case nbar0: return 0.0;
case pplus: return thePDF->xfxQ2(u,x,Q2) - thePDF->xfxQ2(ubar,x,Q2);
default: return 0.0;
}
case ubar:
switch ( particle->id() ) {
case n0: return 0.0;
case pbarminus: return thePDF->xfxQ2(u,x,Q2) - thePDF->xfxQ2(ubar,x,Q2);
case nbar0: return thePDF->xfxQ2(d,x,Q2) - thePDF->xfxQ2(dbar,x,Q2);
case pplus:
default: return 0.0;
}
case d:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(u,x,Q2) - thePDF->xfxQ2(ubar,x,Q2);
case pbarminus: return 0.0;
case nbar0: return 0.0;
case pplus: return thePDF->xfxQ2(d,x,Q2) - thePDF->xfxQ2(dbar,x,Q2);
default: return 0.0;
}
case dbar:
switch ( particle->id() ) {
case n0: return 0.0;
case pbarminus: return thePDF->xfxQ2(d,x,Q2) - thePDF->xfxQ2(dbar,x,Q2);
case nbar0: return thePDF->xfxQ2(u,x,Q2) - thePDF->xfxQ2(ubar,x,Q2);
case pplus:
default: return 0.0;
}
case ParticleID::g:
return 0.0;
}
return 0.0;
}
double ThePEG::LHAPDF::xfsx(tcPDPtr particle, tcPDPtr parton,
Energy2 partonScale, double x,
double, Energy2) const {
// Here we should return the actual density.
using namespace ThePEG::ParticleID;
double Q2 = partonScale / GeV2;
if ( ! thePDF->inRangeXQ2(x, Q2) ) {
switch ( rangeException ) {
case rangeThrow: Throw<Exception>()
<< "Momentum fraction (x=" << x << ") or scale (Q2=" << Q2
<< " GeV^2) was outside of limits in PDF " << name() << "."
<< Exception::eventerror;
break;
case rangeZero:
return 0.0;
case rangeFreeze:
x = min(max(x, xMin), xMax);
Q2 = min(max(Q2, Q2Min/GeV2), Q2Max/GeV2);
}
}
int pid = parton->id();
int abspid = abs(pid);
switch ( pid ) {
case t:
case tbar:
case b:
case bbar:
case c:
case cbar:
return maxFlav() < abspid ? 0.0 : thePDF->xfxQ2(pid,x,Q2);
case s:
case sbar:
return thePDF->xfxQ2(pid,x,Q2);
case u:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(dbar,x,Q2);
case pbarminus: return thePDF->xfxQ2(ubar,x,Q2);
case nbar0: return thePDF->xfxQ2(dbar,x,Q2);
case pplus: return thePDF->xfxQ2(ubar,x,Q2);
default: return thePDF->xfxQ2(u ,x,Q2);
}
case ubar:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(dbar,x,Q2);
case pbarminus: return thePDF->xfxQ2(ubar,x,Q2);
case nbar0: return thePDF->xfxQ2(dbar,x,Q2);
case pplus:
default: return thePDF->xfxQ2(ubar,x,Q2);
}
case d:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(ubar,x,Q2);
case pbarminus: return thePDF->xfxQ2(dbar,x,Q2);
case nbar0: return thePDF->xfxQ2(ubar,x,Q2);
case pplus: return thePDF->xfxQ2(dbar,x,Q2);
default: return thePDF->xfxQ2(d ,x,Q2);
}
case dbar:
switch ( particle->id() ) {
case n0: return thePDF->xfxQ2(ubar,x,Q2);
case pbarminus: return thePDF->xfxQ2(dbar,x,Q2);
case nbar0: return thePDF->xfxQ2(ubar,x,Q2);
case pplus:
default: return thePDF->xfxQ2(dbar,x,Q2);
}
case g:
return thePDF->xfxQ2(g,x,Q2);
case ParticleID::gamma:
- return 0.;
+ return thePDF->xfxQ2(ParticleID::gamma,x,Q2);
}
return 0.0;
}
void ThePEG::LHAPDF::persistentOutput(PersistentOStream & os) const {
os << thePDFName << theMember << theMaxFlav
<< xMin << xMax << ounit(Q2Min, GeV2) << ounit(Q2Max, GeV2);
}
void ThePEG::LHAPDF::persistentInput(PersistentIStream & is, int) {
is >> thePDFName >> theMember >> theMaxFlav
>> xMin >> xMax >> iunit(Q2Min, GeV2) >> iunit(Q2Max, GeV2);
initPDFptr();
}
ThePEG::DescribeClass<ThePEG::LHAPDF, ThePEG::PDFBase>
describeLHAPDF("ThePEG::LHAPDF", "ThePEGLHAPDF.so");
void ThePEG::LHAPDF::Init() {
static ClassDocumentation<LHAPDF> documentation
("The LHAPDF class inherits from PDFBase and implements an interface "
"to the LHAPDF library of parton density function parameterizations. "
"This class is available even if LHAPDF was not properly installed "
"when ThePEG was installed, but will then produce an error in the "
"initialization. Note that the valence densities from the xfvx() and "
"xfvl() function will only work properly for nucleons. All other "
"particles will have zero valence densities.");
static Deleted<LHAPDF> interfacePType
("PType",
- "The LHAPDFv6 interface currently does not support pi or gamma.");
+ "The LHAPDFv6 interface currently does not support pi.");
static Parameter<LHAPDF,string> interfacePDFName
("PDFName",
"The name of the PDF set to be used. Should correspond to "
"the LHAPDF v6 name.",
&ThePEG::LHAPDF::thePDFName, "THEPEG_NO_PDFSET_CHOSEN", true, false,
&ThePEG::LHAPDF::setPDFName);
static Deleted<LHAPDF> interfacePDFNumber
("PDFNumber",
"Not implemented in the LHAPDFv6 interface. "
"Use :PDFName and :Member instead.");
static Parameter<LHAPDF,int> interfaceMember
("Member",
"The chosen member of the selected PDF set.",
&ThePEG::LHAPDF::theMember, 0, 0, 0,
true, false, Interface::lowerlim,
&ThePEG::LHAPDF::setPDFMember);
static Deleted<LHAPDF> interfacePDFLIBNumbers
("PDFLIBNumbers",
"Not implemented in the LHAPDFv6 interface. "
"Use :PDFName and :Member instead.");
- static Deleted<LHAPDF> interfaceEnablePartonicGamma
- ("EnablePartonicGamma",
- "The LHAPDFv6 interface currently does not support pi or gamma.");
-
- static Deleted<LHAPDF> interfacePhotonOption
- ("PhotonOption",
- "The LHAPDFv6 interface currently does not support pi or gamma.");
-
static Deleted<LHAPDF> interfaceMaxNSet
("MaxNSet",
"Not required in LHAPDFv6.");
static Command<LHAPDF> interfaceTest
("Test",
"Write out the values of the chosen PDF set using the x, Q2 and P2 "
"parameters supplied.",
&ThePEG::LHAPDF::doTest, true);
static Deleted<LHAPDF> interfaceVerboseLevel
("VerboseLevel",
"LHAPDFv6 uses general debug level instead.");
static Parameter<LHAPDF,int> interfaceMaxFlav
("MaxFlav",
"The maximum number of flavours for which non-zero densities are "
"reported. The actual number of flavours may be less depending on "
"the chosen PDF set.",
&ThePEG::LHAPDF::theMaxFlav, 5, 3, 0,
true, false, Interface::lowerlim);
interfacePDFName.rank(9);
interfaceMember.rank(8);
}