***************START of README**************************************************
README for RacoonWW : A Monte Carlo generator for
=================== 4-fermion production at e+ e- colliders
********************************************************************************
RacoonWW, Version 1.3.1, released on March 19, 2004
RacoonWW can be downloaded from
https://www.hepforge.org/downloads/racoonww/
********************************************************************************
Summary of changes since Version 1.3.1:
-> We removed a bug in the unweighting routine. This change only affects
the unweighting routine and thus has no impact on any of our published
results, which have been always produced with weighted events.
We are grateful to Fabio Cossutti whose studies with RacoonWW let to
the detection of this bug.
********************************************************************************
Summary of changes in the updated version of RacoonWW 1.3:
06/18/03
--------
-> For smc=3,src=1,sisr=1, higher order ISR, i.e. beyond O(alpha), is now
taken into account in LL approximation up to O(alpha^3), and the
default value for deltas has been changed to deltas=0.001. The need to
include terms of O(alpha^3) occurred in an analysis of the W mass shift
extracted from W invariant mass distributions M(f1,f2), which have
been rescaled with a factor E_beam/(E_f1+E_f2). We thank Fabio
Cossutti for drawing our attention to this effect.
For smc=2,src=0,1,2, a minor bug has been corrected which only affects
the optimization of the a-priori weights.
02/05/03
--------
-> A bug in the calculation of the momenta of the collinear ISR photons
in the common block 'hepevt' has been corrected. This bug concerns
only these momenta in the common block 'hepevt' and does not enter the
calculation of the cross section. Thus, this change does not affect
the numerical results published by the RacoonWW collaboration.
********************************************************************************
Summary of changes since Version 1.2:
-> The off-shell Coulomb singularity has been also implemented for
Born ee->4f (src=0 and sborn4=1).
Accordingly, the flag 'scoul5' has been renamed to 'scoultree'.
-> A minor bug in the ISR structure functions (smc=1,2) has been corrected
(we checked that this change has no significant numerical impact).
-> Photons resulting from collinear initial-state radiation are
included in the HEPEVT common block.
********************************************************************************
Summary of changes since Version 1.1:
-> Anomalous triple gauge-boson couplings to e+e- -> 4f have been
implemented. The charged anomalous triple gauge-boson couplings are
defined as in G. Gounaris et al., in `Physics at LEP2', eds. G. Altarelli,
T. Sj"ostrand and F. Zwirner, (CERN-96/01, Geneva, 1996) p. 79,
hep-ph/9601233. Our conventions for the neutral anomalous triple
gauge-boson couplings differ from those of G.J. Gounaris, J. Layssac and
F.M. Renard, Phys. Rev. D61 (2000) 073013 [hep-ph/9910395] and
Phys. Rev. D62 (2000) 073012, [hep-ph/0005269] only by a minus sign in
the Z-boson field.
-> In the RC mode (src=1) with sisr=1, initial-state radiation in the
leading-log approximation is NOT restricted anymore
to the CC11 class of diagrams, but can be taken into account
to the full process e+e- -> 4f (as it was already the case for src=0).
For src=1 and sisr=1, LL ISR is now calculated with the same options
as the Born cross section, i.e. with ssigepem4, sqcdepem and satgc
according to your choices in the input file.
-> Interface with PYTHIA is available (see Section IV).
-> The different components of the main program 'racoonww.f':
reading the input,
initialization,
event generation,
interface with Pythia, and
writing the output
have been reorganized in form of subroutines for each step.
-> The name of the flag 'qpaw' has changed: qpaw -> spaw
-> The name of the flag 'nevents' has changed: nevents -> neventsw
-> As default, the file 'optimization.info' is not generated.
-> Two more distributions are calculated (dat.26, dat.27).
-> The input files 'inputiba' and 'inputqgc' have been renamed
to 'inputiba4f' and 'inputiba4fa', respectively,
and the input file 'inputtree' has been replaced by 'inputtree4f'.
-> Note that the input files of Version 1.1 are not compatible with
those of Version 1.2.
********************************************************************************
Summary of changes since Version 1.0:
-> Implementation of an IBA for ee->4f+photon.
---------
-> The flag for IBA 'siba' has been removed and instead the
option 'src' has been extended as follows:
src=2: IBA for ee->WW->4f
src=3: IBA for ee->4f+photon
-> Anomalous quartic gauge-boson couplings to ee->4f+photon have been
implemented, including the operators L_0, L_c, L_n (see [4] for details).
-> The input files 'inputdef0' and 'inputdef1' have been renamed to
'inputiba' and 'inputtree', respectively, and a new input file 'inputqgc'
has been added.
-> Note that the input files of Version 1.0 are not compatible with
those of Version 1.1.
********************************************************************************
RacoonWW has been tested under
Operating Systems: UNIX, LINUX
Computers: ALPHA/DEC workstation (Digital Fortran 77)
LINUX PC (Red Hat Linux 6.1,
GNU project Fortran Compiler (v0.5.24))
Language: Fortran 77
********************************************************************************
Authors:
========
A. Denner, University of Wuerzburg, Germany
S. Dittmaier, University of Freiburg, Germany
M. Roth,
D. Wackeroth, SUNY at Buffalo, USA
********************************************************************************
A brief description:
====================
RacoonWW is a Monte Carlo generator for 4-fermion production at
e+ e- colliders.
In the tree-level mode RacoonWW provides TREE-LEVEL cross sections to
- all processes e+ e- -> 4 fermions with anomalous
triple gauge-boson couplings
- all processes e+ e- -> 4 fermions + photon with anomalous
quartic gauge-boson couplings
(* - CC03 or CC11 class of processes e+ e- -> 4 fermions + gluon *)
(* NOT supported by Version 1.3 of RacoonWW! *)
In the Radiative-Correction mode (RC mode) RacoonWW calculates
ELECTROWEAK RADIATIVE CORRECTIONS to
- e+ e- -> W W -> 4f in Double-Pole Approximation (DPA).
When calculating RADIATIVE CORRECTIONS, RacoonWW includes
- the full electroweak O(alpha) corrections to e+ e- -> W W -> 4f:
- complete photon radiation with full off-shell kinematics
- electromagnetic leading-logarithmic virtual corrections with full
off-shell kinematics
- remaining virtual corrections in DPA
- HIGHER-ORDER electroweak corrections:
- higher-order (i.e. beyond O(alpha)) Initial-State Radiation (ISR)
up to O(alpha^3)
via the structure function approach (CERN 96-01, page 129)
- leading effects from \delta\rho and \delta\alpha by using the G_mu
scheme.
- QCD O(alpha_s) corrections:
- either in form of a naive QCD factor attached to each hadronically
decaying W boson or
(* - by a full matrix element calculation of QCD *)
(* O(alpha_s) corrections to e+ e- -> WW -> 4f. *)
(* NOT supported by Version 1.3 of RacoonWW! *)
RacoonWW also provides an IMPROVED BORN APPROXIMATION (IBA) for the calculation
of electroweak radiative corrections to e+ e- -> W W -> 4f and for
e+ e- -> 4f + photon.
The IBA for e+ e- -> W W -> 4f is based on leading universal corrections only
(see [3] for details) and is thus a simple alternative to the full O(alpha)
correction as calculated in the RC mode which is also applicable in the
threshold region. Comparing the IBA with the result of the RC mode illustrates
the size of the non-leading electroweak corrections.
The IBA for e+ e- -> 4f+photon includes LL ISR up to O(alpha^3), where the soft
photons are exponentiated, and the leading weak effects parametrized by G_mu.
In case of WW mediated processes (CC and MC (a),(d)) also the off-shell Coulomb
singularity can be included (see [4] for details).
RacoonWW assumes massless fermions, both in the calculation of the phase space
and the matrix elements.
RacoonWW is a multi-channel Monte Carlo, where the a priori weights are
optimized to improve the accuracy of the integration (R. Kleiss and R. Pittau,
Comp. Phys. Comm. 83 (1994) 141). The a priori weights are the probabilities
for choosing a specific phase-space generator.
RacoonWW generates weighted events. A "hit and miss" unweighting procedure is
provided.
********************************************************************************
Documentation: published in Comput.Phys.Commun.153, 462 (2003), hep-ph/0209330
==============
Publications:
=============
The tree-level predictions and matrix elements for e+ e- -> 4f (+ photon) are
given in
[1] PREDICTIONS FOR ALL PROCESSES E+E- --> 4 FERMIONS + GAMMA,
Nucl. Phys. B560 (1999) 33-65 [hep-ph/9904472]
The implementation of the radiative corrections is described in
[2] ELECTROWEAK RADIATIVE CORRECTIONS TO E+E- --> WW --> 4 FERMIONS
IN DOUBLE POLE APPROXIMATION: THE RACOONWW APPROACH,
Nucl. Phys. B587 (2000) 67-117 [hep-ph/0006307]
The IBA for ee->WW->4f is described in
[3] OFF-SHELL W-PAIR PRODUCTION - UNIVERSAL VERSUS NON-UNIVERSAL CORRECTIONS,
contribution to the Proceedings of the "5th International Symposium on
Radiative Corrections" RADCOR2000, Carmel, September 2000, hep-ph/0101257
The implementation of genuine anomalous quartic gauge-boson couplings
and the IBA for ee->4f+photon are described in
[4] PROBING ANOMALOUS QUARTIC GAUGE-BOSON COUPLINGS VIA
E+E- --> 4 FERMIONS + GAMMA, hep-ph/0104057
Additional numerical results have been published in
[5] O(ALPHA) CORRECTIONS TO E+E- --> WW --> FOUR FERMIONS (+GAMMA):
FIRST NUMERICAL RESULTS FROM RACOONWW,
Phys. Lett. B475 (2000) 127-134 [hep-ph/9912261]
[6] W PAIR PRODUCTION AT FUTURE E+ E- COLLIDERS: PRECISE PREDICTIONS FROM
RACOONWW,
EPJdirect Vol.2 C4 (2000) 1 [hep-ph/9912447]
********************************************************************************
I. Getting started
===============
1. If you have not already done so, please unpack the uuencoded file
RacoonWW.uu:
There are two possibilities, either by executing the shell script
> sh RacoonWW.uu
or by hand
> uudecode RacoonWW.uu
> gunzip RacoonWW.tar.gz
> tar -xvf RacoonWW.tar
2. After the unpacking there are 11 fortran files, a makefile, an inputfile,
a number of default input and output files:
inputtree4f, inputiba4f, inputiba4fa, inputsli, inputsub
outputtree4f, outputiba4f, outputiba4fa, outputsli, outputsub,
and a directory 'data' with the corresponding data-files.
The main program is 'racoonww.f'.
Please edit the makefile and make changes according to your working
environment:
- Please change the FC (=fortran compiler) and FFLAGS (flags for the
compiler) if necessary.
For running on alpha/dec workstations we use
FC=f77
FFLAGS=-O
On a LINUX PC we use the gnu fortran compiler
FC=g77
FFLAGS=-O -fno-emulate-complex -ffast-math
As default, RacoonWW does not need any external libraries.
3. Now you should be able to compile and run RacoonWW.
For compilation enter
> make
and if successful an executable file 'RacoonWW' is created.
For running RacoonWW enter
> RacoonWW < inputfile
To test if everything works fine please use first 'inputiba4f'.
Just enter
> RacoonWW < inputiba4f
This calculates the total cross section to e+ e- -> u d mu n_mu at 200 GeV
using the improved-Born approximation (IBA). Information on the input and
the result for the total cross section are written in 'outputfile'. The
data for the histograms are stored in files dat.01, ..., dat.27. Please
compare the file 'outputfile' with 'outputiba4f' to check if everything
works as it should. In the directory 'data/iba4f' the corresponding
histograms are provided, data/iba4f/dat.01, ..., data/iba4f/dat.27, which
should be compared with dat.01, ..., dat.27. Please find more details
about the input and output of RacoonWW in Section III.
********************************************************************************
II. SLICING and SUBTRACTION branches
================================
RacoonWW is a merger of two independent Monte Carlo programs, which are called
the SLICING and SUBTRACTION branch. The two branches of RacoonWW are based on
the same matrix elements, but the phase-space generation is completely
independent. In the case of radiative corrections, the two branches also
differ in the treatment of the infra-red and collinear singularities. The
slicing branch is based on the phase-space slicing method and the subtraction
branch employs the subtraction method as described in
- S. Dittmaier, Nucl. Phys. B565 (2000) 69-122 [hep-ph/9904440].
- M. Roth, dissertation ETH Zurich No. 13363, 1999 [hep-ph/0008033].
The comparison of the results obtained with the SLICING and SUBTRACTION
branches provides a powerful numerical check of RacoonWW.
Fortran-files used by all branches:
racoonww.f, public.f, ee4fa_amps.f, eeWW4f_DPA.f, eeWW4f_amps.f,
onelints_ext.f, onelints.f, pawgraphs.f
-pawgraphs.f
is only used if you have chosen to use PAW/HBOOK (spaw=1, see below) for
generating histograms in addition to our default histogram routine. Then
you need to link CERNLIB, i.e. please (un)comment the lines
#LIBS= -L/usr/local/cern/99/lib -lmathlib -lkernlib -lpawlib -lpacklib
LIBS=
so that
LIBS= -L/usr/local/cern/99/lib -lmathlib -lkernlib -lpawlib -lpacklib
#LIBS=
appears in the makefile and set the path -L/usr/local/cern/99/lib to the
appropriate path to the CERN libraries on your machine. If the CERN
libraries are not installed on your machine, they can be obtained from
http://wwwinfo.cern.ch/asd/install/Welcome.html.
SLICING branch (smc=1 or 3)
==========================
Fortran-files used only by the slicing branch: slicing.f, kern.f
-slicing.f
SUBROUTINE SLICING
initializes the slicing branch of RacoonWW.
SUBROUTINE ADDMAP
determines the kinematical channels for a given final state.
-kern.f
This is the MC kernel, where the phase space is generated and the matrix
elements are calculated. Here changes are strongly discouraged.
SUBTRACTION branch (smc=2)
==========================
Fortran-files used only by the subtraction branch: subtraction.f
-subtraction.f
SUBROUTINE INITSUBTRACTION
initializes the subtraction branch of RacoonWW.
SUBROUTINE SUBTRACTION
is the MC kernel where the weights and momenta are calculated.
Here changes are strongly discouraged.
********************************************************************************
III. The input and output of RacoonWW
================================
IIIa. INPUT
=====
The input and the different options provided by RacoonWW can be chosen by
modifying the input file 'inputfile' without actually having to look inside the
code. It is advisable to stay within the boundaries given by the input file.
If your choices are not consistent and could lead to wrong results, RacoonWW
prints a warning and either overwrites your choices or stops the program
execution. If your choice is overwritten, the warning contains a reference to
the README, where you can find more information.
In case you would like to change more than what is permitted by the input file,
RacoonWW includes the fortran file 'public.f'. In 'public.f' the values for
the SM input parameters can be changed, additional phase-space cuts can be
defined, and histograms can be added or modified. Options that are not
provided in the input file are explicitly set to their default values in
'racoonww.f'.
RacoonWW comes with five default input files:
- 'inputtree4f' for the calculation of the tree-level process
e+ e- -> u u u u at 190 GeV (see Ref. [1], note however,
that the input parameters differ from those in Ref. [1])).
- 'inputiba4f' for the calculation of the IBA to the process
e+ e- -> u d mu n_mu at 200 GeV.
- 'inputiba4fa' for the calculation of the IBA to the process
e+ e- -> u d mu n_mu + photon at 200 GeV (see Ref. [4], note however,
that the cuts differ from those in Ref. [4]).
- 'inputsli' and 'inputsub' for the calculation of the radiative
corrections to e+ e- -> WW -> u d mu n_mu at 200 GeV, i.e.
our 'best' result (see Ref. [2]), using the slicing and
subtraction branches, respectively.
These input files contain our RECOMMENDED CHOICES for the switches for
the different modes offered by RacoonWW: tree level, IBA and RC mode.
For our 'best' result the recommended switches differ from those used
in the comparisons of the LEP2 Monte Carlo workshop only in the change of
qnf=2 to qnf=3 and ssigepem4=1 to ssigepem4=0. To reproduce the results
of the LEP2 Monte Carlo workshop, we suggest to use the default input files
('inputsli' or 'inputsub') with qnf=2 and ssigepem4=5 (=CC11 Born), and to
subtract the CC11 Born and add the CC03 Born from the result.
We give no recommendation for the branch (smc=1,2,3).
Instead we advise to check the results by comparing the different branches.
The corresponding output files are 'outputtree4f', 'outputiba4f',
'outputiba4fa', 'outputsli', 'outputsub', and the corresponding data files for
the distributions can be found in the directories data/tree4f, data/iba4f,
data/iba4fa, data/sli, data/sub, respectively. The results have been obtained
by running RacoonWW under UNIX on a DEC workstation.
Here is a description of the input file, where it is not self-explanatory.
We use 'inputsli' as an example.
------------------------start-of-inputfile--------------------------------------
outputfile ! name of output file
200d0 ! energy: CMF energy (in GeV)
********************************************************************************
* NOTE: The positron goes into the +z direction.
* NOTE: In the RC mode (src=1), RacoonWW is tested up to an CMF energy of
* 500 GeV. The theoretical uncertainty of the total cross section was
* estimated to be 0.4% for energies between 200 and 500 GeV, 0.5% at 180
* GeV, and 0.7% for 170 GeV (see hep-ph/0005309). Below about 170 GeV the
* DPA is not applicable and above 500 GeV contributions become numerically
* important which are not yet implemented in RacoonWW, e.g. large
* logarithmic electroweak corrections beyond O(alpha). In the tree-level
* mode RacoonWW is tested up to CMF energies of 10 TeV (see [1]) for
* tree-level processes, and IBA results have been obtained for CMF
* energies up to 1 TeV. The tree-level and IBA modes work also in the
* WW threshold region.
********************************************************************************
50000000 ! neventsw: number of weighted events
********************************************************************************
* NOTE: The number of weighted events must be at least neventsw=2000000 in the
* tree-level and IBA modes and neventsw=10000000 in the RC mode to
* guarantee that the multi-channel integration yields reliable estimates
* for the MC integration error.
* When unweighted events are generated (neventsunw>0),
* neventsw is the number of weighted events used to initialize the
* unweighting procedure, i.e. they are used
* to determine the maximal weights (see Section IV for more details).
********************************************************************************
1 ! smc: choice of MC branch: 1(or 3):slicing 2:subtraction
********************************************************************************
* smc=1: the default choice for the slicing branch.
* smc=2: the subtraction branch.
* smc=3: a slightly modified version of the slicing branch to minimize the
* number of negative weights when calculating radiative corrections
* (src=1). For sisr=1 higher-order ISR is included in the leading-log
* approximation up to O(alpha^3). In the RC mode, RacoonWW provides
* an unweighting procedure only for this branch.
* More details can be found in Section IV.
*
* NOTE.1: smc=3 is only a valid choice in the RC mode (src=1). RacoonWW sets
* sborn4=1 in 'racoonww.f' independent of your choice.
********************************************************************************
1 ! sborn4: include Born ee->4f: 0:no 1-3:yes
********************************************************************************
* sborn4=0: the Born contribution ee->4f is NOT included.
* sborn4=1-3: the Born contribution ee->4f is included:
* sborn4=1: off-shell Born contribution ee->4f
* (selection of diagrams ruled by flag ssigepem4, see below)
* sborn4=2: DPA Born, i.e. the Born cross section based on the CC03 diagrams
* to ee->WW->4f. The full off-shell phase space is generated but the
* matrix element is calculated with on-shell-projected momenta.
* (the Z-boson width is set to zero).
* sborn4=3: DPA Born, i.e. the Born cross section based on the CC03 diagrams
* to ee->WW->4f. The phase space and the matrix element are
* calculated with on-shell kinematics.
* (the Z-boson width is set to zero).
*
* NOTE: sborn4=2,3 should only be used for checks and comparisons but not for
* generation of physical results. sborn4=2,3 is only a valid choice for
* the CC, MC(a) and MC(d) class of processes (see below in 'specification
* of final-state fermions' for the definition of CC, MC(a), and MC(d)).
* sborn4=2,3 is also only a valid choice for sisr=0.
* Otherwise RacoonWW stops the execution with a warning.
********************************************************************************
1 ! sborn5: include Born ee->4f+photon: 0:no 1:yes
********************************************************************************
* sborn5=0: the Born contribution ee->4f+photon is NOT included.
* sborn5=1: the Born contribution ee->4f+photon is included.
*
* NOTE: In the RC mode (src=1) and IBA mode (src=3) RacoonWW sets sborn5=1
* independent of your choice. In the slicing branch (smc=1,3), when
* radiative corrections to ee->WW->4f are calculated (src=1), sborn5=1
* calculates the hard photon bremsstrahlung contribution, i.e. photon
* radiation AWAY from soft and collinear singularities.
********************************************************************************
0 ! sborng5: include Born ee->4f+gluon: 0:no 1:yes
********************************************************************************
* sborng5=0: the Born contribution ee->4f+gluon is NOT included.
* sborng5=1: the Born contribution ee->4f+gluon is included.
*
* (* sborng5=1 NOT supported by Version 1.3 of RacoonWW! *)
*
********************************************************************************
1 ! sisr: include higher-order ISR: 0:no 1:yes
********************************************************************************
* sisr=0 : higher-order ISR is NOT included
* sisr=1 : higher-order ISR (in leading-log approximation)
* up to order alpha^3 is included
*
* NOTE: In general, the higher-order ISR contribution in leading-log
* approximation is scale-dependent. For sisr=1 the scale dependence is
* beyond DPA of O(alpha). The scale can be set in 'racoonww.f'. As
* default RacoonWW sets scale=energy.
*
* When calculating radiative corrections (src=1), only the ISR beyond
* O(alpha) is switched on/off by the choice of sisr, i.e. the full
* O(alpha) ISR contribution is always included as part of the explicit
* matrix element calculation even when sisr=0. Note that for src=1,
* sisr=0 the full O(alpha) ISR contribution is included only for the CC11
* class of processes. For sisr=1, the leading-log contribution to the
* ISR is taken into account for the full 4-fermion process, i.e. also for
* the background diagrams. When anomalous triple gauge-boson couplings
* are considered (satgc=1), the complete ISR in LL approximation
* (O(alpha)+higher-orders) to the additional contributions involving
* anomalous couplings is also switched on/off by the choice of sisr.
*
* When calculating IBA (src=2,3) or ISR to ee->4f+photon (src=0, sisr=1,
* sborn5=1), the full ISR is always taken into account in leading-log
* approximation up to O(alpha^3).
********************************************************************************
1 ! src: include radiative corrections: 0:no 1:DPA 2:IBA-4f 3:IBA-4fa
********************************************************************************
* src=0 : radiative corrections are NOT included
* src=1 : radiative corrections are included
* src=2 : Improved-Born Approximation (IBA) for ee->WW->4f
* is calculated for the CC03 class of processes
* src=3 : Improved-Born Approximation (IBA) for ee->4f+photon
* is calculated for the full matrix element.
*
* NOTE: RacoonWW can be run in the tree-level mode (src=0) OR in the RC mode
* (calculation of radiative corrections to ee->WW->4f in DPA) (src=1) OR
* in the IBA mode (src=2) for ee->WW->4f OR in the IBA mode (src=3) for
* ee->4f+photon. In the tree-level mode you can choose between the
* calculation of the tree-level ee->4f processes (sborn4=1) OR the
* calculation of the tree-level processes ee->4f+photon (sborn5=1) OR the
* calculation of the tree-level processes ee->4f+gluon (sborng5=1). In
* the latter two cases and in the IBA mode for ee->4f+photon appropriate
* separation cuts must be applied to avoid IR and collinear singularities
* (see example below).
*
* NOTE: The allowed options for the different modes are summarized in TABLE.1
* of Section E. If an IBA is calculated (src=2,3), only the following
* values for the flags sborn4, sborn5, sborng5, sisr, qalp, qgw, qprop,
* ssigepem4, qqcd are allowed:
*
* IBA for ee->WW->4f (src=2):
*
* sborn4=0
* sborn5=0
* sborng5=0
* sisr=1
* qalp=2
* qgw=0 or 2
* qprop=1
* ssigepem4=1
* qqcd=0 or 2
*
* Here sborn4=0 because a special 4f-matrix element is used for the IBA.
* For qgw=2 the IBA W width is calculated (see below). The choice of
* qalp has no effect on the amplitude and on the width when src=2.
* Moreover, the naive QCD corrections can be included when src=2.
*
* RacoonWW overwrites your choices, if not consistent with src=2.
* RacoonWW sets qqcd=2 in case you chose qqcd=1 or 3 and qgw=2 if you
* choose qgw=1,3,4.
*
* IBA for ee->4f+photon (src=3):
*
* sborn4=0
* sborn5=1
* sborng5=0
* sisr=1
* qalp=2
* qgw=0 or 2
* qprop=1
* qqcd=0 or 2
* sqcdepem=0 or 1
* scoultree=0,1 or 2 for CC and MC (a),(d) processes
* scoultree=0 for NC and MC (b),(c) processes
*
* RacoonWW overwrites your choices, if not consistent with src=3.
* RacoonWW sets qqcd=2 in case you chose qqcd=1 or 3 and qgw=2 if you
* choose qgw=1,3,4.
********************************************************************************
1 ! scoultree: choice of Coulomb singularity for ee->4f,4fA: 0:no 1,2:yes
********************************************************************************
* scoultree=0 : off-shell Coulomb singularity to ee->4f,4fA NOT included
* scoultree=1-2: off-shell Coulomb singularity to ee->4f,4fA included
* scoultree=1 : The Coulomb singularity is calculated with invariant masses
* derived from the four-momenta of the final-state fermions only.
* scoultree=2 : The Coulomb singularity is calculated with those invariant
* masses which are closest to the W mass, i.e. with the possible
* inclusion of the photon momentum. Note that the invariant mass of
* the resonant W boson may or may not include the photon momentum
* depending whether the photon is emitted from the initial or final
* state (see [4]).
*
* NOTE: This option is only effective if src=0(sborn4=1 or sborn5=1) or src=3.
* scoultree=1,2 is only a valid choice for WW mediated processes, i.e. for
* the CC and MC (a),(d) class of processes. RacoonWW sets scoultree=0
* otherwise, independent of your choice. scoultree=2 is forbidden for
* src=0,sborn4=1 and RacoonWW sets scoultree=1 in this case.
********************************************************************************
3 ! qnf: choice of Coulomb singularity in RC mode: 1,2, or 3
********************************************************************************
* qnf=1-3 : different implementations of the Coulomb singularity included
* in the virtual non-factorizable corrections with:
* (The real non-factorizable corrections are implicitly contained
* in the full matrix-element calculation of ee->4f+photon.)
* qnf=1: DPA Coulomb singularity
* qnf=2: off-shell Coulomb singularity with DPA Born cross section
* qnf=3: off-shell Coulomb singularity with (off-shell) CC03 Born cross section
*
* NOTE.2: This option is only effective if src=1.
* Since for src=1 and qnf=0 IR singularities are not properly cancelled,
* qnf=0 is not a valid choice and will be overwritten (qnf=3). If you
* are interested in calculating the DPA part without non-factorizable
* corrections, you have to explicitly set qnf=0 in 'racoonww.f' and the
* result will depend on the photon mass lambda!
*
* NOTE: In [2] different versions of the DPA have been investigated by modifying
* the implementation of the DPA and comparing the results. Different
* treatments of the Coulomb singularity, qnf=2 ('def') and qnf=1 ('Coul')
* is one of the considered options. While the recommended setting at
* the time of the LEP2 Monte Carlo Workshop was qnf=2 we now recommend the
* improved setting qnf=3.
********************************************************************************
1 ! qreal: neglect imaginary part of virt. corr.: 0:no 1:yes
********************************************************************************
* qreal=0: imaginary parts of the 1-loop corrections are included
* qreal=1: imaginary parts of the 1-loop corrections are NOT included
*
* NOTE: This option is only effective if src=1.
*
* NOTE: The imaginary parts of the 1-loop corrections contribute only to
* distributions that depend on the azimuthal decay angle of one of the W
* bosons. If such distributions are considered, qreal has to be set to 0
* (imaginary parts are not neglected). Otherwise the recommended choice is
* qreal=1.
********************************************************************************
2 ! qalp: choice of input-parameter scheme: 0,1, or 2
********************************************************************************
* qalp=0-2 : choice of input-parameter scheme:
* qalp=0 : alpha(0) scheme
* qalp=1 : alpha(MZ) scheme
* qalp=2 : G_mu scheme
*
* NOTE: Independent of the choice of the scheme, the electroweak corrections
* are always calculated by using the fine structure constant alpha(0),
* i.e. the relative (electroweak) corrections
* (dsigma-dsigma_Born)/dsigma_Born are always proportional to alpha(0).
*
* NOTE: Not all input parameters defined in the SUBROUTINE PARAMETER (public.f)
* are used for different input-parameter schemes, i.e. alphaz is only
* relevant for qalp=1, GF only for qalp=2. While the small fermion masses
* have to be non-zero for all input parameter schemes, their actual values
* matter only if qalp=0.
********************************************************************************
4 ! qgw: calculate the W-boson width: 0:no 1-4:yes
********************************************************************************
* qgw=0 : The total W-boson width is an input parameter:
* gw in SUBROUTINE PARAMETER in 'public.f'.
* qgw=1-4: The total W-boson width is calculated:
* qgw=1 : The tree-level W-boson width is calculated.
* qgw=2 : The IBA W-boson width is calculated.
* qgw=3 : The one-loop W-boson width is calculated
* without QCD O(alpha_s) corrections.
* qgw=4 : The one-loop W-boson width is calculated
* with QCD O(alpha_s) corrections.
*
* When the one-loop W-boson width is calculated (qgw=3 or 4 and src=0,1), the
* full electroweak O(alpha) corrections in the chosen input-parameter scheme
* (controlled by flag qalp) are included.
* When qgw=4 also the QCD O(alpha_s) corrections are included.
* When qgw=2 the W width is calculated in IBA, i.e. the tree-level W width in
* the G_mu scheme (qalp=2) including QCD O(alpha_s) corrections.
* Note that qgw=0 or 2 are the only allowed options for src=2,3.
*
* In 'racoonww.f' the partial decay widths for the decays in leptons and quarks
* are also calculated using the same options as for the calculation of the
* amplitude. In the output file, RacoonWW provides the "effective" branching
* ratios: leptonic width/total width, hadronic width/total width and
* (leptonic+hadronic width)/total width. The widths in the numerators are
* calculated with the same options as the DPA matrix element for ee->WW->4f.
* Since the total W width and the partial widths are not necessarily calculated
* with the same options, the total BR can be different from one. The effective
* BRs are only provided for your information, they are NOT used in RacoonWW.
********************************************************************************
1 ! qprop: choice of width scheme
********************************************************************************
* qprop=0-4: choice of width scheme:
* qprop=0: gauge-boson width = 0
* qprop=1: constant width is used in all propagators
* qprop=2: constant width is used only in s-channel (time-like) propagators
* qprop=3: running width is used only in s-channel (time-like) propagators
* qprop=4: constant width is used in all propagators and
* the weak mixing angle is complex
*
* NOTE: If radiative corrections are included (src=1) or for IBA (src=2,3),
* always the constant width approach (qprop=1) is used, i.e. independent
* from your choice RacoonWW sets qprop=1 for src=1,2,3!
********************************************************************************
0 ! ssigepem4: choice of diag. for Born ee->4f: 0:all 1-5:subsets
********************************************************************************
* ssigepem4=0-5: Selection of diagrams for the tree-level 2->4 contribution
* ssigepem4=0: all diagrams
* " 1-4: only signal diagrams (e+e- -> V1V2 -> 4f )
* " 1: V1V2 = WW (CC03)
* " 2: V1V2 = ZZ (NC02)
* " 3: V1V2 = ZZ,gammaZ,Zgamma,gammagamma
* " 4: V1V2 = WW,ZZ,gammaZ,Zgamma,gammagamma
* " 5: only diagrams for CC11 class
*
* NOTE: The choice of ssigepem4 affects the tree-level cross sections to
* ee->4f and, for sisr=1, in addition the leading-logarithmic corrections
* from ISR.
********************************************************************************
5 ! ssigepem5: choice of diag. for Born ee->4f+photon: 0:all 1-5:subsets
********************************************************************************
* ssigepem5=0-5: Selection of diagrams for tree-level 2->5 photon contribution
* ssigepem5= 0: all diagrams
* " 1-4: only signal diagrams (e+e- -> V1V2 -> 4f+photon)
* " 1: V1V2 = WW
* " 2: V1V2 = ZZ
* " 3: V1V2 = ZZ,gammaZ,Zgamma,gammagamma
* " 4: V1V2 = WW,ZZ,gammaZ,Zgamma,gammagamma
* " 5: only diagrams for CC11 class
*
* NOTE: The choice of ssigepem5 only affects the tree-level cross sections
* to ee->4f+photon when src=0,3. When calculating radiative corrections
* (src=1), always the CC11 class of diagrams is used, i.e. RacoonWW
* always sets ssigepem5=5 independent of your choice.
********************************************************************************
0 ! ssigepemg5: choice of diag. for Born ee->4f+gluon: 1:CC03 5:CC11
********************************************************************************
* ssigepemg5=1,5: Selection of diagrams for tree-level 2->5 gluon contribution
* ssigepemg5= 1: only signal diagrams (e+e- -> WW -> 4f+gluon)
* " 5: only diagrams for CC11 class
*
* NOTE.3: The choice of ssigepemg5 only affects the tree-level cross sections to
* ee->4f+gluon when src=0 (qqcd=0). When calculating QCD corrections
* (src=1, qqcd=1 or 3), the choice of diagrams is determined by the
* choice of qqcd (see below), i.e. RacoonWW always sets ssigepemg5=1
* for qqcd=1 and ssigepemg5=5 for qqcd=3 independent of your choice.
********************************************************************************
2 ! qqcd: include QCD radiative corr.: 0:no 1:CC03 2:naive 3:CC11
********************************************************************************
* qqcd=0 : QCD corrections are NOT included.
* qqcd=1 : QCD corrections are included via a matrix-element calculation
* for ee->WW->4f based on the CC03 class of diagrams.
* qqcd=2 : 'naive' QCD corrections are included, i.e. for each hadronically
* decaying W boson RacoonWW applies a global factor (1+alpha_s/pi)
* to the differential cross sections.
* qqcd=3 : QCD corrections are included via a matrix-element calculation for
* ee->4f based on the CC11 class of diagrams.
*
* NOTE: qqcd=1 or 3 is not a valid choice when src=0! Please use qqcd=2 when
* you would like to include (naive) QCD corrections to the tree-level
* processes ee->4f and ee->4f+photon. When the tree-level processes
* ee->4f+gluon are calculated (sborng5=1,src=0) only qqcd=0 is allowed.
*
* (* qqcd=1,3 NOT supported by Version 1.3 of RacoonWW! *)
*
********************************************************************************
0 ! sqcdepem: include gluon-exch. diagrams in Born: 0:no 1:yes 2:only
********************************************************************************
* sqcdepem=0 : gluon-exchange background diagrams are NOT included.
* sqcdepem=1 : gluon-exchange background diagrams are included.
* sqcdepem=2 : ONLY gluon-exchange background diagrams are included.
* (no purely electroweak ones)
*
* NOTE: The choice of sqcdepem only affects the tree-level cross sections to
* ee->4f and ee->4f+photon, and for isr=1 also the leading-logarithmic
* corrections from ISR. This QCD contribution then consists of
* gluon-exchange background diagrams to the tree-level processes ee->4f
* and ee->4f+photon as described in Ref. [1]. In the RC mode (src=1)
* gluon-exchange diagrams are taken into account only in the tree-level
* matrix element ee->4f.
********************************************************************************
u ! fermion 3
d ! anti-fermion 4
mu ! fermion 5
nu_mu ! anti-fermion 6
********************************************************************************
* specification of final-state fermions:
*
* fname(3:6): u,d,s,c,t,b,e,mu,tau,nu_e,nu_mu,nu_tau
*
* Please observe the following order:
* (f,F =/= e-,n_e, f',F' = isospin partners of f,F, resp.)
*
* anti-f1 f2 --> f3 anti-f4 f5 anti-f6
* CC processes:
* e+ e- --> f anti-f' F anti-F'
* e+ e- --> n_e e+ f anti-f'
* e+ e- --> f anti-f' e- anti-n_e
* NC processes:
* e+ e- --> f anti-f F anti-F
* e+ e- --> f anti-f f anti-f
* e+ e- --> e- e+ f anti-f
* e+ e- --> e- e+ e- e+
* mixed CC-NC processes (=MC processes):
* (a) e+ e- --> f anti-f f' anti-f'
* (b) e+ e- --> n_e anti-n_e f anti-f
* (c) e+ e- --> n_e anti-n_e n_e anti-n_e
* (d) e+ e- --> n_e anti-n_e e- e+
*
* If you choose another order for the final state, RacoonWW stops with a
* warning.
*
********************************************************************************
0d0 ! pp: degree of positron beam polarization [-1d0:1d0]
0d0 ! pm: degree of electron beam polarization [-1d0:1d0]
********************************************************************************
1 ! srecomb: photon recombination: 0:no 1:TH 2:EXP 3:TH(no cuts)
********************************************************************************
* srecomb=0 : no recombination is applied
* srecomb=1 : with TH recombination scheme (see also [2]):
* The minimal invariant mass of the photon and any charged
* final-state fermion, M(photon,fcomb), is determined, and the
* photon is recombined with the fermion fcomb when
* M(photon,fcomb) < precomb(3) or
* the photon energy E(photon) < precomb(2).
* srecomb=2 : with EXP recombination scheme:
* The minimal angle between the photon and any charged final-state
* fermion, theta(photon,fcomb), is determined, and the photon is
* recombined with the fermion fcomb when
* theta(photon,fcomb) < precomb(3) in case fcomb is a lepton or
* theta(photon,fcomb) < precomb(4) in case fcomb is a quark or
* the photon energy E(photon) < precomb(2).
* srecomb=3 : with TH recombination scheme and recombination with the beams:
* The minimal invariant mass of the photon and any charged fermion
* (including the incoming electron and positron), M(photon,fcomb),
* is determined, and the photon is recombined with the fermion fcomb
* when it is a final-state fermion and when
* M(photon,fcomb) < precomb(3) or
* the photon energy E(photon) < precomb(2).
* If fcomb=beam electron/positron, only the photon momentum is set
* to zero.
*
* NOTE: When srecomb=1,2,3 the parameters precomb(1:4) must be specified in
* the input file. RacoonWW stops the execution with a warning if one of
* the parameters is set to zero! For srecomb=3 no separation cuts are
* required and precomb(1)=0 can be chosen.
*
* WARNING: If the unweighting procedure is used, the recombination procedure
* and separation cuts influence the resulting unweighted events.
*
* The recombination scheme is defined in SUBROUTINE RECOMBINATION in 'public.f'.
* The TH recombination scheme (srecomb=1) is described in detail in Ref. [2]
* (with precomb(1)=5d0, precomb(2)=1d0, precomb(3)=M_rec=5 GeV (bare) or 25 GeV
* (calo)).
********************************************************************************
5d0 ! precomb(1): angular rec. cut between photon and beam
********************************************************************************
* precomb(1) : srecomb=1,2,3: separation cut on angle between beam
* and photon (in degrees)
*
* The photon momentum is set to zero, when the angle between the photon and the
* beam is smaller than precomb(1). This cut is applied at the very beginning
* before photon recombination is performed.
*
* NOTE: precomb(1) is not active when srecomb=0.
********************************************************************************
1d0 ! precomb(2): rec. cut on photon energy
********************************************************************************
* precomb(2) : srecomb=1,2,3: recombination cut on photon energy (in GeV)
*
* The photon is recombined with a charged final-state fermion (as explained
* above) when its energy is smaller than precomb(2).
*
* NOTE: precomb(2) is not active when srecomb=0.
********************************************************************************
5d0 ! precomb(3): inv.-mass rec.(TH) or angular rec. for lept.(EXP)
********************************************************************************
* precomb(3) : srecomb=1,3: invariant mass recombination cut (in GeV)
* srecomb=2: angular recombination cut for leptons (in degrees)
*
* NOTE: precomb(3) is not active when srecomb=0.
********************************************************************************
0d0 ! precomb(4): angular rec. cut for quarks(EXP)
********************************************************************************
* precomb(4) : srecomb=1,3: N/A
* srecomb=2: angular recombination cut for quarks (in degrees)
*
* NOTE: precomb(4) is not active when srecomb=0, 1, or 3.
********************************************************************************
0 ! srecombg: gluon recombination scheme: 0:no 1:TH 2:EXP
0d0 ! precombg(1): rec. cut on gluon energy
0d0 ! precombg(2): inv.-mass (TH) or angular (EXP) recombination cut
********************************************************************************
* srecombg=0 : no recombination is applied
*
* (* values of srecombg other than 0 *)
* (* NOT supported by Version 1.3 of RacoonWW! *)
*
********************************************************************************
0 ! satgc: anomalous triple gauge-boson couplings (ATGCs): 0:no 1:yes
********************************************************************************
* satgc=0: anomalous triple gauge-boson couplings are NOT included.
* satgc=1: anomalous triple gauge-boson couplings are included.
* The parameters:
* Delta g_1^(A,Z), Delta kappa^(A,Z), lambda^(A,Z),
* tilde kappa^(A,Z), tilde lambda^(A,Z), (g,f,h)_i^(A,Z)
* must be specified in the input file (see below).
*
* For the definition of the charged anomalous triple gauge-boson
* couplings we use the conventions of G. Gounaris et al., in
* `Physics at LEP2', eds. G. Altarelli, T. Sj"ostrand and F. Zwirner,
* (CERN-96/01, Geneva, 1996) p. 79, hep-ph/9601233. Our conventions
* for the neutral anomalous triple gauge-boson couplings differ from
* those of G.J. Gounaris, J. Layssac and F.M. Renard, Phys. Rev. D61
* (2000) 073013 [hep-ph/9910395] and Phys. Rev. D62 (2000) 073012,
* [hep-ph/0005269] only in a minus sign in the Z-boson field.
*
* NOTE: The anomalous triple gauge-boson couplings only affect the
* tree-level ee->4f cross sections and, for sisr=1, also the
* leading-logarithmic ISR corrections are applied to the tree-level cross
* sections involving anomalous triple gauge-boson couplings.
********************************************************************************
0d0 ! ATGC Delta g_1^A
0d0 ! ATGC Delta g_1^Z
0d0 ! ATGC Delta kappa^A
0d0 ! ATGC Delta kappa^Z
0d0 ! ATGC lambda^A
0d0 ! ATGC lambda^Z
0d0 ! ATGC g_4^A
0d0 ! ATGC g_4^Z
0d0 ! ATGC g_5^A
0d0 ! ATGC g_5^Z
0d0 ! ATGC tilde kappa^A
0d0 ! ATGC tilde kappa^Z
0d0 ! ATGC tilde lambda^A
0d0 ! ATGC tilde lambda^Z
0d0 ! ATGC f_4^A
0d0 ! ATGC f_4^Z
0d0 ! ATGC f_5^A
0d0 ! ATGC f_5^Z
0d0 ! ATGC h_1^A
0d0 ! ATGC h_1^Z
0d0 ! ATGC h_3^A
0d0 ! ATGC h_3^Z
********************************************************************************
0 ! qaqgc: anomalous quartic gauge-boson couplings (AQGCs): 0:no 1:yes
********************************************************************************
* qaqgc=0: anomalous quartic gauge-boson couplings are NOT included.
* qaqgc=1: anomalous quartic gauge-boson couplings are included.
* The parameters a_i/Lambda^2
* must be specified in the input file (see below). For the definition
* of the anomalous quartic gauge-boson couplings we refer to Ref. [4].
*
* NOTE: The inclusion of anomalous quartic gauge-boson couplings is not
* supported in the RC mode (src=1).
* The anomalous couplings only affect the results obtained in
* the IBA mode (src=3) and the tree-level mode (src=0,sborn5=1) for
* ee->4f+photon. RacoonWW sets qaqgc=0 for src=1 independent of your
* choice.
********************************************************************************
0d0 ! AQGC a_0/Lambda^2
0d0 ! AQGC a_c/Lambda^2
0d0 ! AQGC a_n/Lambda^2
0d0 ! AQGC tilde a_0/Lambda^2
0d0 ! AQGC tilde a_n/Lambda^2
********************************************************************************
10 ! scuts: separation cuts
********************************************************************************
* scuts=0,1,2,10,11 : choice of separation cuts
* scuts=0 : NO separation cuts are applied
* scuts=1 : ADLO cuts are applied as defined in racoonww.f
* scuts=2 : LC cuts are applied as defined in racoonww.f
* scuts=10 : a minimal set of cuts must be specified in the input file
* (all energies and invariant masses in GeV, all angles in degrees):
***************************************************************************
0d0 ! photon energy cut
0d0 ! charged-lepton energy cut
0d0 ! quark energy cut
0d0 ! quark-quark invariant mass cut
0d0 ! angular cut between photon and beam
0d0 ! angular cut between photon and charged lepton
0d0 ! angular cut between photon and quark
0d0 ! angular cut between charged leptons
0d0 ! angular cut between quarks
0d0 ! angular cut between charged lepton and quark
10d0 ! angular cut between charged lepton and beam
10d0 ! angular cut between quark and beam
********************************************************************************
* scuts=11 : cuts on the energies of all final-state particles, on all angles
* and invariant masses must be specified in the input-file:
********************************************************************************
*0d0 ! energy cut on particle 3
*0d0 ! energy cut on particle 4
*0d0 ! energy cut on particle 5
*0d0 ! energy cut on particle 6
*0d0 ! energy cut on photon
*0d0 ! angular cut between beam1 and particle 3
*0d0 ! angular cut between beam1 and particle 4
*0d0 ! angular cut between beam1 and particle 5
*0d0 ! angular cut between beam1 and particle 6
*0d0 ! angular cut between beam1 and photon
*0d0 ! angular cut between beam2 and particle 3
*0d0 ! angular cut between beam2 and particle 4
*0d0 ! angular cut between beam2 and particle 5
*0d0 ! angular cut between beam2 and particle 6
*0d0 ! angular cut between beam2 and photon
*0d0 ! angular cut between particle 3 and 4
*0d0 ! angular cut between particle 3 and 5
*0d0 ! angular cut between particle 3 and 6
*0d0 ! angular cut between particle 3 and photon
*0d0 ! angular cut between particle 4 and 5
*0d0 ! angular cut between particle 4 and 6
*0d0 ! angular cut between particle 4 and photon
*0d0 ! angular cut between particle 5 and 6
*0d0 ! angular cut between particle 5 and photon
*0d0 ! angular cut between particle 6 and photon
*0d0 ! particle 3-particle 4 invariant-mass cut
*0d0 ! particle 3-particle 5 invariant-mass cut
*0d0 ! particle 3-particle 6 invariant-mass cut
*0d0 ! particle 3-photon invariant-mass cut
*0d0 ! particle 4-particle 5 invariant-mass cut
*0d0 ! particle 4-particle 6 invariant-mass cut
*0d0 ! particle 4-photon invariant-mass cut
*0d0 ! particle 5-particle 6 invariant-mass cut
*0d0 ! particle 5-photon invariant-mass cut
*0d0 ! particle 6-photon invariant-mass cut
********************************************************************************
* NOTE: Non-standard cuts, i.e. cuts which are not covered by the choices of
* scuts that RacoonWW provides, can be included in SUBROUTINE CUT in
* 'public.f'.
********************************************************************************
------------------------end of inputfile----------------------------------------
A. Recommended choices of input-flags in the tree-level mode: src=0
================================================================
A.1. inputtree4f: calculation of tree-level processes ee->4f
========================================================
We recommend to use the tree-level mode (src=0) for processes ee->4f that
do not include W-pair production, e.g. when calculating cross sections
for NC processes.
While some CC processes (without final-state electrons/positrons) can be
calculated without separation cuts (scuts=0), appropriate cuts must be applied
for NC and MC processes to avoid collinear singularities. We advise to use
the ADLO (scuts=1) and LC (scuts=2) set of cuts for LEP2 and LC energies,
respectively.
Using 'inputtree4f' RacoonWW calculates the Born cross sections to the NC
process e+ e- -> u u u u at 190 GeV for massless, unpolarized fermions.
All diagrams, pure electroweak and gluon-exchange diagrams, are included, the
tree-level W width is calculated, the G_mu scheme and constant width are used,
ADLO cuts are applied, and the subtraction branch of RacoonWW is used. No
anomalous triple gauge-boson couplings are included.
NOTE: To obtain results that correspond to the input of [1] (Table 1) please
make the following modifications:
-edit 'public.f' and set the parameters alpha_s, M_W, M_Z, gamma_W,
gamma_Z and G_mu in SUBROUTINE PARAMETER to the values of [1] (Sec.4.1).
-edit 'racoonww.f' and calculate the mixing angle cw by using the formula
given in Ref. [1] (Sec.4.1) in SUBROUTINE INITIALIZE.
-choose qgw=0, qalp=0 and set alpha0=1d0/128.07d0 in SUBROUTINE
PARAMETER.
A.2. Calculation of tree-level processes ee->4f+photon
==================================================
In the tree-level mode (src=0) also hard-photon processes ee->4f+photon can be
studied. To do this, the inputfile of Section A.1 can be used with the
modifications sborn4=0, sborn5=1.
NOTE: To obtain results that correspond to the input of [1] (Table 1) please
make the modifications described in the NOTE in Section A.1.
Note that appropriate cuts must be applied for all processes to avoid IR and
collinear singularities. We advise to use scuts=1 (ADLO) for LEP2 energies
and scuts=2 (LC) for LC energies. If you would like to define your own cuts,
use scuts=10 (minimal set of cuts) or scuts=11.
NOTE: scuts=0 is NOT allowed! RacoonWW will print a warning (and stop the
program execution) if no cuts are applied.
B. Recommended choices of input-flags in the IBA mode: src=2,3
===========================================================
B.1. inputiba4f: IBA for ee->WW->4f
==============================
This calculates the CC03 cross section to e+ e- -> W W -> u d mu n_mu in IBA
at 200 GeV including leading-log ISR up to O(alpha^3) and 'naive' QCD. The
total W-boson width is calculated in IBA. No cuts are applied. No anomalous
triple gauge-boson couplings are included. The subtraction branch of RacoonWW
is used.
NOTE: The choice src=2 is only possible for CC, MC(a) and MC(d) processes.
For all other processes RacoonWW stops the execution with a warning when
src=2!
B.2. inputiba4fa: IBA for ee->4f+photon
==================================
This contains the recommended choices of the input-flags when calculating
the hard-photon processes ee->4f+photon.
It calculates the cross section to e+ e- -> u d mu n_mu + photon in IBA at
200 GeV including leading-log ISR up to O(alpha^3) and 'naive' QCD. The total
W-boson width is calculated in IBA. ADLO cuts are applied, and the subtraction
branch of RacoonWW is used. No anomalous quartic gauge-boson couplings are
included.
C. Recommended choices of input-flags in the RC mode: src=1
=========================================================
We recommend to use the RC mode (src=1) for the investigation of W-pair
production processes ee->WW->4f(+photon).
Electroweak radiative corrections are calculated. The real radiation is
restricted to the CC11 class of diagrams and the non-leading virtual
corrections are calculated in DPA. However, the leading-log ISR contribution
up to O(alpha^3) to the full 4-fermion process is included.
NOTE: NC, MC(b) and MC(c) processes (see above in 'specification of final-state
fermions' for the definition of NC, MC(b), and MC(c)), i.e. processes
without intermediate doubly-resonant W bosons, are NOT allowed! RacoonWW
will print a warning (and stop the program execution), if an
inappropriate final state is chosen.
C.1. inputsub, inputsli: full radiative corrections to ee->WW->4f in DPA
===================================================================
This calculates the full electroweak corrections (in the G_mu scheme) to e+ e-
-> W W -> u d mu n_mu in DPA at 200 GeV including the full Born cross section
and higher-order ISR, when using the subtraction branch (inputsub) or the
slicing branch (inputsli). QCD O(alpha_s) corrections are included by using a
'naive' QCD factor. The total W-boson width is calculated including
electroweak and QCD one-loop corrections.
NOTE: This is the default setup for the calculation of our 'best' results as
presented in the LEP2 MC workshop report (hep-ph/0005309) and in Ref. [2]
(please see remarks in the beginning of Section IIIa concerning changes
in RacoonWW since the LEP2 MC workshop).
A recombination scheme (TH 'bare') as described in Ref. [2] and separation cuts
on the angle between the beams and the charged final-state fermions are
applied. No other cuts are applied.
When calculating radiative corrections to ee->WW->4f the ee->4f (2->4) and
ee->4f+gamma (2->5) events are calculated simultaneously.
The 2->4 events include the virtual corrections in DPA, the soft and collinear
parts of real photon emission, the Born contribution (sborn4=1) and
higher-order ISR (sisr=1). In the subtraction branch the 2->5 events
correspond to the real process ee->4f+photon and the subtraction functions of
the 5-particle phase space. In the slicing branch the 2->5 events describe
the hard bremsstrahlung contribution, i.e. real photon radiation away from the
soft and collinear singularities.
In the slicing branch (smc=1(or 3)) the hard bremsstrahlung contribution is
obtained by imposing technical cuts on the photon energy (deltas) and on the
cosine of the angle (deltac) between the photon and any charged final-state
fermion. Consequently, the 2->4 and 2->5 parts depend on deltas and deltac;
only in the sum the dependence on the technical cuts cancels. Please find
more details on the technical cuts in section D.2.
In the subtraction branch (smc=2), the real corrections are obtained by
subtracting suitable subtraction functions from the matrix element squared of
the bremsstrahlung process over the full phase space, so that the integral
contains no singularities. The subtracted singular contributions are added
back after performing the singular integrations over the photonic part of the
phase space analytically with regulators for the infrared and collinear
singularities.
If you are only interested in the total cross section, separation cuts and
recombination are not necessary (scuts=0,srecomb=0). Please recall that
photon recombination itself only redistributes events in distributions.
However, for distributions a recombination scheme srecomb=1, 2, or 3 must be
applied. If you modify the recombination scheme, you have to make sure that
it results in IR- and collinear-safe observables. To this end, we advise to
carefully compare the results obtained with the SLICING and SUBTRACTION
branches; strong differences between these branches typically signal non-IR-
or non-collinear-safe observables.
********************************************************************************
D. Options not provided in the input file (expert mode!)
=====================================================
RacoonWW has a few more options, which are not provided in the input file.
Changing those options, is only advisable for 'experts', i.e. for those who are
familiar with the details of our calculation as described in Ref. [2]. If they
are common to all MC branches, RacoonWW sets them to their default values in
'racoonww.f' (after 'setting options'). Special options for the slicing and
subtraction branches are set in SUBROUTINE INITSLICING and SUBROUTINE
INITSUBTRACTION, respectively.
D.1. Options common to all branches
==============================
The following flags are set in 'racoonww.f' in SUBROUTINE INITIALIZE:
son, qfast, qsoft, spaw
The first three flags only affect the calculation of the virtual corrections,
i.e. only the DPA part.
- flag 'son'
=========
son=1,2,3,4: choice of on-shell projection for virtual corrections in DPA.
RacoonWW provides four different versions for the projection of off-shell
momenta to on-shell momenta for
e^+(p1)+e^-(p2) -> f1(p3)+ fbar2(p4) + f3(p5)+ fbar4(p6):
( CMF of e^+e^- is assumed )
son = 1: directions of momenta of W^+, f1 and f3 are fixed (default!)
son = 2: directions of momenta of W^+, f2 and f3 are fixed
son = 3: directions of momenta of W^+, f1 and f4 are fixed
son = 4: directions of momenta of W^+, f2 and f4 are fixed
Since the projections are performed in the CMF, the direction of the W^- is
opposite to the direction of the W^+.
The projection is done in SUBROUTINE OFTOON in 'eeWW4f_DPA.f'.
NOTE: In [2] the accuracy of the DPA has been investigated by modifying the
implementation of the DPA and comparing the results. The on-shell
projections, son=1 ('def') and son=2 ('proj'), is one of the considered
options.
- flag 'qfast'
============
qfast=0: form factors for virtual corrections from full formulas (slow!)
qfast=1: expansion into Legendre polynomials (fast!) (default!)
Details of this approach can be found in Section 3.1.2 of [2].
- flag 'qsoft'
============
qsoft=1,2,3: definition of the DPA part of radiative corrections, terms
subtracted from the virtual corrections in DPA are fixed by:
qsoft=1: endpoint part as defined in Eq. (4.29) of [2] (default!)
qsoft=2: only logarithms as defined in Eq. (4.2.53) of hep-ph/0008033
qsoft=3: YFS factor for virtual photons as defined in Eq. (4.31) of [2]
NOTE: Results obtained with different choices for qsoft differ only by terms
that are beyond DPA accuracy (see [2]).
NOTE: In [2] the accuracy of the DPA has been investigated by modifying the
implementation of the DPA and comparing the results. The definition of
the finite virtual corrections, qsoft=1 ('def') and qsoft=3 ('eik'), is
one of the considered options.
- flag 'spaw'
==========
RacoonWW provides the possibility to use HBOOK/PAW for histogram filling and
plotting. The usage of HBOOK/PAW (spaw=1) is only allowed when generating
weighted events (neventsunw=0). If you would like to use it in addition to our
default histogram routine, please do the following:
- Go to 'racoonww.f': set the switch spaw=1 and uncomment all lines in the
if-environment starting with
if(spaw.eq.1 ...
- Edit 'makefile' and add 'pawgraphs.o' to the list of OBJS.
In 'pawgraphs.f' the histograms are defined in SUBROUTINE SETUP_GRAPHS and
filled in SUBROUTINE GRAPHS (GRAPHS_B, GRAPHS_K and GRAPHS_R). This is the
place, where you could add and/or modify histograms.
NOTE: When using HBOOK/PAW, CERNLIB must be linked
(see remarks in Sec.II concerning the makefile)!
D.2. Options specific for the slicing branch (smc=1 (or 3))
======================================================
The following flags and parameters are set in SUBROUTINE INITSLICING:
dsdc, dsdcg, opt, nopt, iopt, noptg, ioptg, nu
- flags 'dsdc' and 'dsdcg'
=======================
In the slicing branch (smc=1) the hard photonic bremsstrahlung contribution is
obtained by imposing technical cuts on the photon energy
E_gamma > deltas*E_beam
and on the cosine of the angle between the photon and any charged final-state
fermion
-1 < cos(gamma,f) < 1-deltac
The technical cuts deltas and deltac are set in SUBROUTINE INITSLICING in
'slicing.f'. Consequently, the 2->4 and 2->5 parts depend on deltas and
deltac; only in the sum the dependence on the technical cuts cancels.
When using PAW/HBOOK (spaw=1) the independence of the total cross section
on the technical cuts can be checked as follows:
- Go to SUBROUTINE INITSLICING in 'slicing.f' and set dsdc=1 for the
variation of deltas or dsdc=2 for the variation of deltac.
- Choose the range of variation: loglow, logup
with loglow < log_10(deltas) < logup or loglow < log_10(deltac) < logup
When dsdc=1(2) is chosen, only the histograms number 101(sum), 102(only 2->4,
Born subtracted) and 103(only 2->5) are filled, showing the variation of the
total cross section with the cutoff parameters deltas(deltac). For LEP2
energies we advise to stay within the boundaries of the default values for
loglow and logup as provided in SUBROUTINE INITSLICING, since they correspond
to a plateau as described in Ref. [2] (Figure 4).
When the QCD corrections are calculated (qqcd=1 or 3) the singularities
resulting from to soft and collinear gluon emission are treated in complete
analogy to photon radiation. The 2->4 and 2->5 parts depend on the technical
cuts deltags and deltagc, where
E_gluon > deltags*E_beam
-1 < cos(gluon,quark) < 1-deltagc
The cutoff independence of the total cross section can be checked in the same
way as in the photon case by choosing dsdcg=1(2).
NOTE: Only dsdc=1(2) OR dsdcg=1(2) is allowed.
- flag and parameters for the adaptive optimization:
==================================================
opt=1 : self-optimization of a priori weights
opt=0 : no optimization
The number of events used for the optimization of the a priori weights and how
often the optimization is performed can be separately specified for the
calculation of the 2->4 weights (nopt,iopt) and the 2->5 weights (noptg,ioptg).
The default setting is
nopt=1000000
iopt=10
noptg=25000000
ioptg=1
The flag 'opt' and the parameters enter in FUNCTION KERN in 'kern.f'. The a
priori weights are optimized as long as the number of generated events is
smaller than nopt (or noptg). For the 2->4 part the a priori weights are
optimized after 5000, 10000, 15000, 25000 generated events and when
mod(nevents,int(nopt/iopt))=0. For the 2->5 part the a priori weights are
optimized after 500000, 1500000, 3000000 generated events and when
mod(nevents,int(noptg/ioptg))=0.
- parameter 'nu'
==============
The parameter 'nu' specifies the mapping for the propagators of massless
particles as described in Ref. [1] (Eq. (3.6)). 'nu' enters in SUBROUTINE SMAP
and SUBROUTINE TMAP in 'kern.f'. We strongly advise to use only values not too
different from 1d0, e.g. nu=0.8d0 is a valid choice. The default value is
nu=1d0.
D.3. Options specific for the subtraction branch (smc=2)
===================================================
- flag 'ssub'
===========
In SUBROUTINE INITSUBTRACTION in 'subtraction.f' the flag 'ssub' is set:
ssub=1,2: definition of subtraction functions
ssub=1: definition of
S. Dittmaier, Nucl. Phys. B565 (2000) 69-122 [hep-ph/9904440].
ssub=2: definition of
M. Roth, dissertation ETH Zurich No. 13363, 1999 [hep-ph/0008033].
The results obtained by the two subtraction variants differ only by terms that
are beyond the DPA accuracy. More details can be found in Ref. [2].
- parameters for the adaptive optimization
========================================
nopt(1-8,ng): 'nopt' defines the events for which the adaptive optimization is
performed, i.e. for which event the a priori weights are
recalculated. The optimization steps are given by
nopt(i,ng)*(number of channels), where 'ng' denotes the
phase-space generator (ng=1,2,3) listed below. The adaptive
optimization stops after the i-th optimization step if
nopt(i+1,ng) < nopt(i,ng).
The default settings are
nopt(1,ng)=100
nopt(2,ng)=200
nopt(3,ng)=300
nopt(4,ng)=400
nopt(5,ng)=500
nopt(6,ng)=600
nopt(7,ng)=700
nopt(8,ng)=800
alphamin: The a priori weights have a minimal value which is fixed by
'alphamin'.
- phase-space generators
======================
The phase-space generators are initialized in the SUBROUTINE INITSUBTRACTION
by calling INITPHASESPACE.
Three phase-space generators are used:
1. Four-particle phase-space generator for tree-level process ee->4f,
higher-order LL ISR, and IBA for ee->WW->4f
2. Four-particle phase-space generator for 2->4 part of O(alpha) radiative
corrections in DPA
3. Five-particle phase-space generator for 2->5 part of radiative corrections
and tree-level process or IBA for ee->4f+photon
********************************************************************************
E. Summary of flags
================
Before we turn to the description of the output, all possible values of
(relevant) flags are summarized in TABLE.1:
TABLE.1:
-------------------------------------------------------------------------------
name of | set in | switch for | possible values |
flag | | | ( * means not relevant ) |
-------------------------------------------------------------------------------
src |input file | tree-l./RC mode | src=0 |src=2|src=3|src=1|
-------------------------------------------------------------------------------
| | | tree-level | IBA | IBA | RCs |
| | |ee-> |ee-> |ee-> |ee-> |ee-> | |
| | |4f |4f+ga|4f+gl|4f |4f+ga| |
-------------------------------------------------------------------------------
smc |input file |MC branch | 1,2 | 1,2 | 1,2 | 1,2 | 1,2 | 1-3 |
sborn4 |input file |tree-level ee->4f| 1-3 | 0 | 0 | 0 | 0 | 0-3 |
sborn5 |input file |tree-level 4f+ga | 0 | 1 | 0 | 0 | 1 | 1 |
sborng5 |input file |tree-level 4f+gl | 0 | 0 | 0 | 0 | 0 | 0 |
sisr |input file |higher-order ISR | 0,1 | 0,1 | 0 | 1 | 1 | 0,1 |
scoultree |input file |Coulomb sing. for| 0,1 | 0-2 | 0 | * | 0-2 | * |
| |ee->4f,4f+gamma | | | | | | |
qnf |input file |Coulomb sing. for| * | * | * | * | * | 1-3 |
| |ee->4f (RC) | | | | | | |
qreal |input file |imaginary part | * | * | * | * | * | 0,1 |
qalp |input file |parameter scheme | 0-2 | 0-2 | 0-2 | 2 | 2 | 0-2 |
qgw |input file |W width | 0-4 | 0-4 | 0-4 | 0,2 | 0,2 | 0-4 |
qprop |input file |width approach | 0-4 | 0-4 | 0-4 | 1 | 1 | 1 |
ssigepem4 |input file |sub-diagrams (4f)| 0-5 | * | * | 1 | * | 0-5 |
ssigepem5 |input file |sub-diag.(4f+ga) | * | 0-5 | * | * | 0-5 | 5 |
ssigepemg5|input file |sub-diag.(4f+gl) | * | * | 1,5 | * | * | 1,5 |
qqcd |input file |QCD RCs | 0,2 | 0,2 | 0 | 0,2 | 0,2 | 0,2 |
sqcdepem |input file |gluon-ex. backg. | 0-2 | 0-2 | * | * | 0,1 | 0,1 |
srecomb |input file |photon recomb. | * | 0-3 | * | * | 0-3 | 0-3 |
srecombg |input file |gluon recomb. | * | * | 0 | * | * | 0 |
scuts |input file |sep. cuts | 0-11| 0-11| 0-11| 0-11| 0-11| 0-11|
satgc |input file |ATGCs | 0,1 | * | * | 0,1 | * | 0,1 |
qaqgc |input file |AQGCs | * | 0,1 | * | * | 0,1 | 0 |
qfast |racoonww.f |Legendre exp. | * | * | * | * | * | 0,1 |
son |racoonww.f |on-shell proj. | 1-4 | * | * | * | * | 1-4 |
qsoft |racoonww.f |photon DPA part | * | * | * | * | * | 1-3 |
dsdc |slicing.f |technical cuts | * | * | * | * | * | 0-2 |
dsdcg |slicing.f |technical cuts | * | * | * | * | * | 0-2 |
opt |slicing.f |optimization | 0,1 | 0,1 | 0,1 | 0,1 | 0,1 | 0,1 |
ssub |subtraction.f|subtr. functions | * | * | * | * | * | 1,2 |
-------------------------------------------------------------------------------
********************************************************************************
IIIb. OUTPUT
======
The output consists of a file 'outputfile', or the name you have chosen in the
inputfile, and 27 (or more, if you added your own
histograms) data-files, which contain the data for the distributions. If you
choose spaw=1, there is also a paw-file 'pawplot.paw' with the histograms
generated with HBOOK. If you set iout=/=0 in 'racoonww.f'
a file 'optimization.info' will be generated (default:iout=0).
The weights for each event together with the corresponding four-momenta are
provided in the common block 'event' in 'racoonww.f':
common/event/weight,pbeam,p,weightborn,pborn,nsubevent
nsubevent : number of weights
weight(29): smc=1(3): 8 weights per event
smc=2 : 29 weights per event
(see 1. 'outputfile' below)
p(i,j,k) : momenta after possible ISR,
i.e. p(1,0,k)+p(2,0,k) <= CMF energy
i=1:7 numerates the particles (7=photon),
j=0:3 energy,x,y,z component,
k=1:29 numerates the weights,
pbeam(i,j): beam momenta (before ISR),
i.e. pbeam(1,0)+pbeam(2,0) = CMF energy
i=1,2 numerates the beam,
j=0:3 energy,x,y,z component,
weightborn: weight for tree-level four-fermion process
pborn(i,j): momenta corresponding to weightborn
i=1:7 numerates the particles
j=0:3 energy,x,y,z component
NOTE: The momenta of the initial-state particles after ISR, p(1-2,j,k), are NO
physically measurable quantities.
1. 'outputfile'
============
In the output file RacoonWW provides the following information:
- the input parameters, the effective branching ratios and the choices for
the options,
- the total cross section together with the maximal value of the weight (in
the subtraction branch also the corresponding event and the channel is
given),
- the total cross sections for the subcontributions together with the maximal
values of the weights.
subcontributions for the slicing branch (smc=1):
-----------------------------------------------
weightborn: tree-level four-fermion cross section (=0 for sborn4=0)
Note: weight(1) is defined as follows:
sisr=0: weight(1)=weightborn
sisr=1: weight(1)=0
weight(1) is not explicitely given in the outputfile.
weight(2): 2->4 DPA part
weight(3): 2->4 virtual+soft+final-state collinear photon part
(to avoid double counting LL O(alpha) ISR
is subtracted when sisr=1)
weight(4): 2->4 initial-state collinear photon part (radiation from e+)
(to avoid double counting LL O(alpha) ISR
is subtracted when sisr=1)
weight(5): 2->4 initial-state collinear photon part (radiation from e-)
(to avoid double counting LL O(alpha) ISR
is subtracted when sisr=1)
weight(6)-weightborn:
src=1: 2->4 LL initial-state radiation up to O(alpha^3):
the tree-level process is convoluted with the
structure functions (the Born contribution is
subtracted).
src=2: IBA for ee->WW->4f: the CC03 IBA cross section is
convoluted with the structure functions
weight(7): src=3: IBA for ee->4f+photon
src=1: 2->5 photon bremsstrahlung contribution
src=0: tree-level process ee->4f+photon
weight(8): src=1 (qqcd=1 or 3): 2->5 gluon bremsstrahlung contribution
src=0: tree-level process ee->4f+gluon
subcontributions for the subtraction branch (smc=2):
---------------------------------------------------
weightborn: tree-level four-fermion cross section (=0 for sborn4=0)
weight(1): sisr=0: tree-level four-fermion cross section
sisr=1: vanishes
weight(2): sisr=0: vanishes
sisr=1,siba=0: initial-state radiation up to O(alpha^3)
including tree-level cross section
sisr=1,siba=1: Improved Born Approximation for ee->WW->4f
including CC03-Born
weight(3)+weight(4)+weight(5):
2->4 part of radiative corrections including
src=1: virtual O(alpha) corrections and subtraction functions
src=1,sisr=1: O(alpha) LL ISR omitted to avoid double counting
when both radiative corrections and initial-state
radiation are included
weight(3): contribution with 4f kinematics
weight(4): one photon radiation off e+
weight(5): one photon radiation off e-
weight(6): src=1: 2->5 photonic part of radiative corrections
src=0,sborn5=1: tree-level process ee -> 4f+photon
src=3: Improved Born Approximation for ee->4f+photon
weight(7): src=1,qqcd=1,3: 2->5 gluonic part of radiative corrections
src=0,sborng5=1: tree-level process ee -> 4f+gluon
weight(8-29): src=1: contributions from the subtraction functions
subcontributions for the slicing branch (smc=3):
-----------------------------------------------
weight(1): 2->4 tree-level+
DPA part+virtual+soft+final-state collinear photon part+
O(alpha^3) leading-log ISR when sisr=1 (soft part)
weight(2)=0
weight(3)=0
weight(4): 2->4 initial-state collinear photon part (radiation from e+)
+O(alpha^3) leading-log ISR when sisr=1 (coll. part)
weight(5): 2->4 initial-state collinear photon part (radiation from e-)
+O(alpha^3) leading-log ISR when sisr=1 (coll. part)
weight(6): O(alpha^3) leading-log ISR when sisr=1 (coll. part)
weight(7): 2->5 photon bremsstrahlung contribution
weight(8)=0 (only 'naive' QCD is included for smc=3 and qqcd=2)
Finally, the number of rejected events in per cent is given. Events are
discarded owing to the applied separation cuts and in case of the slicing
branch also owing to the imposed technical cuts. RacoonWW provides the number
of rejected events in the calculation of each weight separately.
2. 'optimization.info'
==================
If iout=/=0 the file 'optimization.info' is generated (default:iout=0).
This file contains information about the kinematical channels used in the
multi-channel integration and about the optimization of the a priori weights
(see also [1]).
2a. slicing branch
==============
First, information on the kinematical channels used for the calculation of the
2->4 and 2->5 events is provided:
All channels for the multi-channel integration of the ee->4f and ee->4f+photon
processes are constructed from two and three basic topologies, respectively.
They are parametrized in terms of the parameters ii(1:5), mapi(1:6) (for the
2->4 part), mapig(1:8) (for the QED 2->5 part), mapiq(1:8) (for the QCD 2->5
part), and vzi(1:3).
The generators for the four- and five-particle phase spaces are described in
terms of these topologies in SUBROUTINE PHSPGEN_4 and SUBROUTINE PHSPGEN_5 in
'kern.f', respectively. The specifications of these parameters for each
channel for a given final state is done in SUBROUTINE ADDMAP in 'slicing.f':
ii(1:5) specify the final-state particles,
mapi(6)=1,2,3 numerates the three basic topologies,
mapi(1:5), mapig(1:5), mapig(7:8), mapiq(1:5), and mapiq(7:8) specify the
mappings for the involved propagators:
mapi=0: no mapping
mapi=1: mapping for massless particles
mapi=2,3(W,Z): mapping for massive particles (Breit-Wigner),
and vzi(1:3) specify the permutations needed to account for all topologies.
In the file 'optimization.info' the parameters mapi(1:6), mapig(1:8),
mapiq(1:8) (if qqcd=1 or 3) for each channel are given.
The calculations of the weights(1:8) use each their own optimization of the a
priori weights for the different channels within the weights(1:8). After each
optimization the information on the a priori weights is provided for each
contributing weight. Moreover, the number of events used in the optimization,
the integral estimate and its statistical error are given.
2b. subtraction branch
==================
The file contains information of the phase-space generators. The list includes
the channel, topology, external particles, and virtual particles of each
phase-space generator (=channel). Please find the 15 topologies in the
SUBROUTINE PHASESPACE (subtraction.f), from which all phase-space generators
are built up by choosing the external particles denoted by ep1, ..., ep7 and
the virtual particles denoted by ip1, ..., ip4. Moreover, for the additional
phase-space generators corresponding to the different subtraction functions
the emitter(em) and spectator(sp) are also given in 'optimization.info'.
The second part includes for each phase-space generator the events for which
the adaptive optimization is performed and the a priori weights in per cent
for each channel.
3. data-files for the distributions
================================
The data for the distributions are written in data-files dat.01, ..., dat.27.
The histograms are defined in SUBROUTINE SETTINGS in 'public.f':
dat.01 contains the total cross section (for test purposes).
data-file distribution in the
dat.02 invariant mass of particle 3 and particle 4
(in case of CC processes this corresponds to
the W+ invariant-mass), [75:85] GeV, 50 bins
dat.03 invariant mass of particle 5 and particle 6
(in case of CC processes this corresponds to
the W- invariant-mass), [75:85] GeV, 50 bins
dat.04 cosine of the angle between particle V1(V1->3+4)
and beam1(=positive z-axis), [-1:1], 50 bins
dat.05 cosine of the angle between particle V2(V2->5+6)
and beam2(=negative z-axis), [-1:1], 50 bins
dat.06 cosine of the angle between particle V1(V1->3+4)
and particle 3, [-1:1], 50 bins
dat.07 cosine of the angle between particle V1(V1->3+4)
and particle 4, [-1:1], 50 bins
dat.08 cosine of the angle between particle V2(V2->5+6)
and particle 5, [-1:1], 50 bins
dat.09 cosine of the angle between particle V2(V2->5+6)
and particle 6, [-1:1], 50 bins
dat.10 energy of the photon, [0:50] GeV, 50 bins
(2->4 weights are added to the 1. bin (E_gamma=0d0)
dat.11 cosine of the angle between photon and beam1(=positive
z-axis), [-1,1], 50 bins (2->4 weights are added to
the last bin (cos\theta=1))
dat.12 minimal angle between photon and charged final-state
fermion, [0:180] degrees, 50 bins (2->4 weights are
added to the 1. bin)
dat.13 energy of particle 3, [0:energy/2] GeV, 50 bins
dat.14 energy of particle 4, [0:energy/2] GeV, 50 bins
dat.15 energy of particle 5, [0:energy/2] GeV, 50 bins
dat.16 energy of particle 6, [0:energy/2] GeV, 50 bins
dat.17 cosine of the angle between particle V1(V1->3+4) and
particle 3 in the rest frame of particle V1, [-1:1], 50 bins
dat.18 cosine of the angle between particle V1(V1->3+4) and
particle 4 in the rest frame of particle V1, [-1:1], 50 bins
dat.19 cosine of the angle between particle V2(V2->5+6) and
particle 5 in the rest frame of particle V2, [-1:1], 50 bins
dat.20 cosine of the angle between particle V2(V2->5+6) and
particle 6 in the rest frame of particle V2, [-1:1], 50 bins
dat.21 angle between decay planes of particles 3,4 and 5,6,
[0:360] degrees, 50 bins
dat.22 angle between decay planes of particles V1,beam1 and V1,4,
[0:360] degrees, 50 bins
dat.23 angle between decay planes of particles V2,beam2 and V2,5,
[0:360] degrees, 50 bins
dat.24 cosine of the angle between particle 3 and 4, [-1:1],
50 bins
dat.25 cosine of the angle between particle 5 and 6, [-1:1],
50 bins
dat.26 cosine of the angle between particle V1(V1->3+4)
and beam1(=positive z-axis) in the CMF of particles V1 and V2,
[-1:1], 50 bins
dat.27 cosine of the angle between particle V2(V2->5+6)
and beam2(=negative z-axis), in the CMF of particles V1 and V2,
[-1:1], 50 bins
The angles of the histograms 21 and 22 are defined in Eq. (7.4) and (7.3) of
Ref. [2], respectively.
The data-files contain the differential cross sections for the chosen setup
and the corresponding tree-level cross sections for ee->4f:
column1 column2 column3 column4 column5
Observable(O) dsigma/dO stat. error tree-level dsigma/dO stat. error
Of course the last two entries are zero when sborn4=0.
In SUBROUTINE SETTINGS in 'public.f' you can define your own histograms as
described at the end of SUBROUTINE SETTINGS.
In addition, for spaw=1 the HBOOK file 'pawplot.paw' is generated. It contains
the histograms for the distributions as specified in 'pawgraphs.f'. In
'racoonww.f' the SUBROUTINE SETUP_GRAPHS is called where the histograms are
defined (call hbook). The SUBROUTINE GRAPHS is called in 'racoonww.f' after
the weights have been calculated. In SUBROUTINE GRAPHS the observables are
defined and the corresponding histograms are filled (call hfill). In addition
to SUBROUTINE GRAPHS three more SUBROUTINEs, GRAPHS_B, GRAPHS_K and GRAPHS_R,
are provided which generate histograms showing the distributions for the
tree-level ee->4f contribution, the 2->4 and 2->5 parts only. For viewing
just enter, for instance,
> paw++
which should open the PAW++ main browser, and load the file 'pawplot.paw'
(File -> Open Hbook file...).
IV. Miscellaneous
=================
Random number generator
=======================
If you like to use another random number generator, go to SUBROUTINE RANS in
'racoonww.f' and replace the body of the subroutine with your random number
generator. Please use only double-precision random number generators
especially for high-statistic runs.
Unweighting procedure
=====================
For src=1 and smc=3, or src=0,2,3, RacoonWW can generate unweighted events.
As default, RacoonWW generates weighted events.
The unweighting procedure is activated as follows:
- The number of generated unweighted events you want to generate must be set
in the main program in 'racoonww.f':
neventsunw: number of generated unweighted events
(default:neventsunw=0: NO unweighted events are generated)
In the RC mode (src=1), to obtain a precision comparable
to our 'best' result, we recommend to use neventsunw=50000.
With an efficiency of 1 unweighted event per 1000 weighted events,
this corresponds to the calculation of approximately
500000000 weighted events.
In the tree-level and IBA modes the efficiency is about a factor 100 better
and we recommend to use neventsunw=500000. Note that the efficiency can
strongly depend on the separation cuts for certain processes.
- The number of weighted events (neventsw) that are used only to determine the
maximal weights for the 'hit and miss' algorithm is set in the
input file. This should be much smaller than what is usually
used without unweighting, for instance, we recommend neventsw=100000.
NOTE: When including radiative corrections (src=1), only the branch smc=3
can be used in the unweighting procedure. Otherwise RacoonWW stops
the execution with a warning.
The unweighted events are generated from weighted events with
a "hit and miss" algorithm. More precisely, for
each weighted event an unweighted event is generated if the corresponding
weight is larger than the product of a random number r (0(random number)*(maximal weight).
Otherwise no unweighted event is generated. The maximal weight (weighttotmax)
is calculated by multiplying the maximal weight of the first neventsw events
with the factor 2:
weighttotmax = 2*(maximal weight of the first neventsw events)
'neventsw' and the factor 2 can be tuned to suppress weights that are larger
than 'weighttotmax' in the main calculation. A warning is written in the
output file if a weight exceeds the maximal weight (weight>weighttotmax) and
'weighttotmax' is redefined by the value of the weight.
The maximal weight of the 2 -> 5 bremsstrahlung contribution exceeds by far the
maximal weights of the other subcontributions. Therefore the other weights are
not calculated for each event in order to enhance the efficiency of the
unweighting procedure. They are calculated only all 'nweight' times but
multiplied by 'nweight'. In this way the probability of an unweighted event
from the different subcontributions remains unchanged. The integers
'nweight(i1)' are chosen in such a way that each subcontribution leads to
roughly the same maximal weight, which has to be smaller than 'weighttotmax'.
The information about the (unweighted) event is provided in
the common block HEPEVT. The four momenta corresponding to the
unweighted events are denoted by phep(5,nmxhep)
(phep(4,nxmhep)=energy, phep(1-3,nxmhep)=x,y,z component
and phep(5,nmxhep)=mass(=0)):
beam (e+): phep(1:5,1)
beam (e-): phep(1:5,2)
final-state fermion 3: phep(1:5,3)
final-state anti-fermion 4: phep(1:5,4)
final-state fermion 5: phep(1:5,5)
final-state anti-fermion 6: phep(1:5,6)
photon: phep(1:5,7)
photon: phep(1:5,8)
photon: phep(1:5,9)
They are stored in the common block HEPEVT.
Note that the positron goes in +z direction and that cuts and photon
recombination influence the resulting unweighted events. The last up to two
photons which are collinear to the beam result from initial-state radiation.
Although negative weights appear, they have only a negligible influence on the
observables that we have studied so far. Nevertheless, RacoonWW also unweights
the events with negative weights (weight=-1). The resulting unweighted
negative-weight events can, for instance, be used to estimate the uncertainty
resulting from discarding the negative-weight events.
In the output file RacoonWW provides the number of unweighted events, the
number of unweighted negative events (weight=-1), and the number of events that
exceed the maximal weight (weight>weighttotmax).
WARNING: Note that the result for the total cross section and for the
distributions are calculated from the weighted events.
Interface with PYTHIA
=====================
To include parton shower and hadronization for colored final states,
RacoonWW provides an interface with PYTHIA for unweighted events
(neventsunw>0) by calling the PYTHIA subroutine SUBROUTINE PY4FRM.
As default, this interface is not active in RacoonWW.
It can be activated as follows:
- link PYTHIA by adding the file with the PYTHIA code, e.g.
pythia1568.f (Version 6.1), to the list of OBJS in the 'makefile'.
PYTHIA can be downloaded from http://www.thep.lu.se/~torbjorn/Pythia.html.
- edit 'racoonww.f' and uncomment the line with the call
of SUBROUTINE PY4FRM in the main program.
Note, that the compilation of PYTHIA needs extensive memory.
As default, we switched off photon radiation in the parton shower (irad=0) to
avoid double counting when radiative corrections are included (src=1). In the
tree-level and IBA modes irad=1 may be chosen. Also tau decays are not
included (itau=0), since in RacoonWW the final-state fermions are considered
to be massless. However, we provide the SUBROUTINE MASSIVE in 'racoonww.f'
which generates a massive two-particle phase space. This could be used to
generate massive kinematics for the tau before interfacing with PYTHIA.
Acknowledgement: We thank Luca Malgeri, Stefan Roth and Arno Straessner
for the incentive to interface RacoonWW to Pythia and for comparing
RacoonWW with YFSWW after hadronization.
********************************************************************************
Finally, please do not hesitate to contact us in case you have questions.
The Racoonians,
Ansgar Denner, ansgar.denner@psi.ch
Stefan Dittmaier, dittmair@mppmu.mpg.de
Markus Roth, roth@mppmu.mpg.de
Doreen Wackeroth, dow@ubpheno.physics.buffalo.edu
***************End of README****************************************************