001
002 herwig_i/herwig_v6202_cdf.doc
003 -----------------------------
004 16-Aug-2001, 02-Oct-2001.
005 (first version 16-Aug-2001,
006 revised 02-Oct-2001, taking into account generator reorganization
007 and use of cdfGen)
008
009
010 Here are some items concerning HERWIG, version 6.2 as set up for CDF
011 run 2. We actually use v 6.202, which differs very little from v 6.201,
012 the main difference being in rearrangements in the common block.
013 For a nice 91-page description of herwig, as of v 6.200, see the official
014 write-up in hep-ph/0011363, G. Corella et al. That write-up can be reached
015 from the Herwig web site, http://hepwww.rl.ac.uk/theory/seymour/herwig/ .
016 Also reachable from that web site is the v 6.2 announcement and a list
017 of changes between v 6.1 and v 6.2.
018 That write-up includes (pages 33 - 37) a list of all the hard sub-
019 processes that are currently available.
020
021 The pure herwig code is in /cdf/code/cdfprod/herwig/v6_202/IRIX+6.5/src
022 /herwig62 .
023 Version 6.2 was released in November 2000. The main changes from version
024 6.1 are added Higgs production processes (both new SM and SUSY) and bug
025 fixes.
026
027
028 Jeremy Lys.
029
030
031 -----------------------------------------------------------------------------
032 -----------------------------------------------------------------------------
033
034 Current status (02-Oct-2001)
035 -----------------------------
036
037
038 By itself, herwig runs and produces sensible output.
039 With cdfSim, herwig runs and produces sensible output.
040 See section below on random numbers and reproducibility.
041
042 On fcdfsgi2, for t.tbar events with underlying event (Process 1706)
043 Herwig takes 0.2 cpu sec/event. Simulation on these events takes
044 ~ 15 cpu sec/event. For SUSY events, 2 partons to 2 spartons (Process
045 3010), the analogous times are 0.7 and ~15 cpu sec/event.
046
047 Herwig produces sensible output for SUSY processes. For SUSY, a file
048 must be read in that gives SUSY particle properties. Examples of such
049 files exist in http://hepwww.rl.ac.uk/theory/seymour/herwig , for
050 standard LHC SUGRA and GMSB points. These files can be produced using
051 Isajet and a new package, HWISSP (combined package ISAWIG, see:
052 www-thphys.physics.ox.ac.uk/users/PeterRichardson/HERWIG/isawig.html).
053 Our aim is to set that up on fcdfsgi2 soon.
054
055 Version 4.09 of StdHep is appropriate for Herwig 6.2, as long as
056 processes involving gravitons are not used. A promised update of Stdhep
057 will include gravitons in its list of particles.
058
059 -----------------------------------------------------------------------------
060 -----------------------------------------------------------------------------
061
062 Some properties of Herwig to be aware of
063 ----------------------------------------
064
065 1. Searching *up* the generator level chain to find ancestors of a particle
066 can give wrong answers and infinite loops. Problem occurs amongst the
067 clusters and quarks and diquarks. So the stdhep library routine
068 stdparentlst.F is not good for herwig. An alternative, which searches
069 *down* the chain, is in cdfsga ~lys/backup/stdparentlst.F .
070 But note that this alternative has problems with isajet and pythia events.
071
072 2. Herwig gives non-zero vhep values (i.e., production vertex x y z t) for
073 most quarks, diquarks, gluons, clusters, etc. (as explained in the
074 documentation). Hence a ttbar event may have 395 entries and 385 displaced
075 vertices. We could get rid of "unwanted" vertices by resetting vhep values
076 to zero (in hwufne.f or hwanal.f, for example).
077
078 3. Herwig has a non-zero lifetime for pizero. So gammas from pizero decay
079 get a different production vertex than the pizero.
080
081 4. In top quark decays in v 6.2 (as in v 6.1 but different from v 5.9)
082 there is an explicit W. That is, we get t -> b W, with W -> (u,dbar) or
083 (c,sbar) or (e/mu/tau,neutrino). But this W is "added later" (see hwudat.f,
084 which lists only 3-body top decays, and hwdhob.f), and its decay is not
085 changed by MODBOS settings.
086
087 -----------------------------------------------------------------------------
088 -----------------------------------------------------------------------------
089
090 To run HERWIG
091 -------------
092 One can run either cdfGen or cdfSim. These exist in code releases
093 (in bin/$BFARCH) and often in development. They are created via
094 gmake generatorMods or gmake SimulationMods resp. Only cdfSim includes
095 possible simulation in the detector.
096 Some example tcl files exist in $CDFSOFT2_DIR/herwig_i/examples. There,
097 herwig_top.tcl and herwig_test.tcl are intended for cdfGen running.
098 And hersim0_test.tcl and hersim1_test.tcl for cdfSim running without
099 and with detector simulation, respectively.
100
101 So one can do something like:
102
103 $CDFSOFT2_DIR/bin/$BFARCH/cdfGen < herwig_test.tcl > herwig_test.log
104
105
106 For a SUSY process, one would have in the tcl file, inside the
107 Herwig talkto, something like:
108
109 Process set 3010
110 Read_susy
111 File_Readsusy set /cdf/scratch/lys/run2aug14/sugra_pt1.1103
112 exit
113
114
115
116 The type of reaction is specified by the Process number, which is set in
117 the talkto (see below).
118 For a list of the possible Process numbers, see the official write-up
119 (hep-ph/0011363, G. Corella et al.), pages 33-53.
120 A few Process numbers are:
121 1453 q q' -> W -> tau.nutau,
122 1500 QCD 2->2 hard parton scattering,
123 1706 t.tbar production,
124 2100 W + jet production,
125 2300 SM Higgs + jet production.
126 If 10000 is added (e.g., 11706) the underlying event is suppressed.
127
128
129 NOTE: For more and better wisdom on tcl files, see Ken Bloom's CDF note
130 5294, and the tcl files in $CDFSOFT2_DIR/SimulationMods/test .
131
132 -------------------------------------------------------------------------------
133 -------------------------------------------------------------------------------
134
135 HERWIG talk-to
136 --------------
137
138 Simple Herwig talk-to example:
139
140
141 talk Herwig
142 help // This gives ~50 lines of information
143 show // ~100 lines, gives values of all the variables
144 Process set 11706 // Note order, "set" does NOT come first
145 call_qq set T // Use QQ to decay B and C hadrons (choices below)
146 qq
147 Decay_Bc set T // Note: default QQ (in cdf interface) is to NOT
148 Decay_Upsilon set T // do these decays; the default in Herwig (in cdf
149 Decay_B_Baryons set T // interface) is also to NOT do these decays.
150 Decay_prompt_charm set T // Seems more sensible for QQ to do them.
151 show
152 exit
153 Masses
154 top set 175. // Decimal point not required (of course)
155 show // Gives values of variables in Menu 11 - Masses
156 exit
157 Prints
158 maxpr set 1 // Gives internal Herwig print of 1 event
159 prvtx set 0 // In Herwig is Boolean, default true, here is integer.
160 exit
161 exit
162
163 -------------------------------------------------------------------------------
164
165 Herwig talk-to details:
166
167 The Herwig manual for version 6 (hep-ph/0011363, G Corcella et al., see
168 above) lists three sets of variables that must or may be set by a user.
169 These are reproduced below, with the Herwig description for each of the
170 three lists in quotes.
171
172 Appended, on the right hand side of the lists, is HERWIG talk-to
173 information, as set up for CDF users in generatorMods/HerwigModule.cc
174 (and ditto.hh).
175 The talk-to "name" is the word required in the talk-to interaction,
176 after giving the menu command. The menu commands are in the Menu table
177 below the lists. G stands for general menu, for which no menu command is
178 needed.
179
180 Default values in the lists below (except the *-ed ones) are as set in
181 the initializing HERWIG routines. Almost all are set in hwigin.f. Some
182 values are set or reset in HerwigInterface.F, and indicated by a star (*)
183 in the default column below.
184 The talk-to variables are preset in HerwigModule, but all except three
185 are reset by Herwig. The three not reset are noshower, nohadronization
186 and File_Readsusy; the first two are used in HerwigModule.cc but not in any
187 Herwig routines (are not in Herwig.inc), and the third is passed as an
188 argument to cdfreadsusy.
189
190 Some of the parameters listed are not (i.e., not currently) accessible in
191 the talk-to. For such parameters there are no entries to the right of the
192 herwig information.
193
194 Beneath the lists of variables and parameters is a list of 5 further
195 variables that it may be useful to change sometimes and that can be set in
196 the talk-to.
197
198 -------------------------------------------
199
200 "As indicated above, a number of variables must be set in the main
201 program to specify what is to be simulated:"
202
203 +----------+------------------------------+-----------+ talk-to
204 | Name | Description | Default |Menu "name"
205 +----------+------------------------------+-----------+
206 | PART1 | Type of particle in beam 1 |*'P '| 01 Beam1
207 | PART2 | Type of particle in beam 2 |*'PBAR '| 01 Beam2
208 | PBEAM1 | Momentum of beam 1 |*900. | 01 Pbeam1
209 | PBEAM2 | Momentum of beam 2 |*900. | 01 Pbeam2
210 | IPROC | Type of process to generate |*1500 | G Process
211 | MAXEV | Number of events to generate |*100 |
212
213 +----------+------------------------------+-----------+
214
215
216
217 "The quantities that may be regarded as adjustable parameters are"
218
219 +----------+----------------------------------+-------+ talk-to
220 | Name | Description |Default|Menu "name"
221 +----------+----------------------------------+-------+
222 | QCDLAM | QCD Lambda (see below) | 0.18 | G Lambda_QCD
223
224 +----------+----------------------------------+-------+
225 | RMASS(1) | Down quark mass | 0.32 | 11 down
226 | RMASS(2) | Up quark mass | 0.32 | 11 up
227 | RMASS(3) | Strange quark mass | 0.50 | 11 strange
228 | RMASS(4) | Charmed quark mass | 1.55 | 11 charm
229 | RMASS(5) | Bottom quark mass | 4.95 | 11 bottom
230 | RMASS(6) | Top quark mass | 174.3 | 11 top
231
232 +----------+----------------------------------+-------+
233 | RMASS(13)| Gluon effective mass | 0.75 | 11 gluon
234
235 +----------+----------------------------------+-------+
236 | VQCUT | Quark virtuality cutoff (added to| 0.48 | 05 Qcutoff
237 | | quark masses in parton showers) | |
238 | VGCUT | Gluon virtuality cutoff (added to| 0.10 | 05 Gcutoff
239 | | effective mass in parton showers)| |
240 | VPCUT | Photon virtuality cutoff | 0.40 | 05 Pcutoff
241
242 +----------+----------------------------------+-------+
243 | CLMAX | Maximum cluster mass parameter | 3.35 | 06 Clmax
244 | CLPOW | Power in maximum cluster mass | 2.00 | 06 Clpow
245 | PSPLT(1) | Split cluster spectrum, non-b | 1.00 | 06 Psplt1
246 | PSPLT(2) | Split cluster spectrum, b | 1.00 | 06 Psplt2
247
248 +----------+----------------------------------+-------+
249 | QDIQK | Maximum scale for gluon->diquarks| 0.00 | 06 Qdiqk
250 | PDIQK | Gluon->diquarks rate parameter | 5.00 | 06 Pdiqk
251
252 +----------+----------------------------------+-------+
253 | QSPAC | Cutoff for spacelike evolution | 2.50 | 05 Spacelike_evolution
254 | PTRMS | Intrinsic pt in incoming hadrons| 0.00 | 05 Pt_incoming_hadrons
255
256 +----------+------------------------------+-----------+
257
258
259
260 "A number of quantities can be reset to control the program and
261 various options:"
262
263 +----------+----------------------------------+-------+ talk-to
264 | Name | Description |Default|Menu "name"
265 +----------+----------------------------------+-------+
266 | NEVHEP | Current no. of events | 0 |
267 | NHEP | Current no. entries in /HEPEVT/ | 0 |
268
269 +----------+----------------------------------+-------+
270 | IPRINT | Printout option |*1 | 14 iprint
271 | MAXPR | Number of events to print out |*0 | 14 maxpr
272 | PRVTX | Include vertex info in print out | .TRUE.| 14 prvtx
273 | NPRFMT | Print to screen or log file | 1 | 14 nprfmt
274 | PRNDEC | Use decimal/hexadecimal in print | .TRUE.| 14 prndef
275 | PRNDEF | Print to screen or log file | .TRUE.| 14 prndef
276 | PRNTEX | Print to latex file |.FALSE.| 14 prntex
277 | PRNWEB | Print to html file |.FALSE.| 14 prnweb
278
279 +----------+----------------------------------+-------+
280 | MAXER | Max number of errors |*100 | G Maxer
281
282 +----------+----------------------------------+-------+
283 | LWEVT | Unit for writing output events |*0 |
284
285 +----------+----------------------------------+-------+
286 | LRSUD | Unit for reading Sudakov table | 0 |
287 | LWSUD | Unit for writing Sudakov table | 77 |
288 | SUDORD | Alpha_s order in Sudakov table |*2 | 02 Sudakov_order
289 | INTER | Order of interp., Sudakov tables | 3 |
290
291 +----------+----------------------------------+-------+
292 | NRN(1) | Random number seed 1 | 17673 | G Randseed1
293 | NRN(2) | Random number seed 2 | 63565 | G Randseed2
294 | WGTMAX | Max weight (0 to search for it) | 0. | 09 Maxwt
295 | NOWGT | Generate unweighted events | .TRUE.| 09 No_weights
296 | AVWGT | Mean event weight | 1.0 |
297 | EFFMIN | Minimum acceptable efficiency | 0.001 | 09 Effmin
298
299 +----------+----------------------------------+-------+
300 | AZSOFT | Soft gluon azimuthal correlations| .TRUE.| 05 Azsoft
301 | AZSPIN | Gluon spin azimuthal correlations| .TRUE.| 05 Azspin
302
303 +----------+----------------------------------+-------+
304 | HARDME | Use hard matrix-el corrections | .TRUE.|
305 | SOFTME | Use soft matrix-el corrections | .TRUE.|
306 | GCUTME | Gluon energy cut in top ME corr. | 2.0 |
307
308 +----------+----------------------------------+-------+
309 | NCOLO | Number of colours | 3 | 10 Colors
310 | NFLAV | Number of (producible) flavours | 6 | 10 Flavors
311
312 +----------+----------------------------------+-------+
313 | MODPDF(I)| PDFLIB structure function set and|*41 | 03 Modpdf
314 | AUTPDF(I)| author group for beam I(=1,2) | 'MRS' | 03 Autpdf
315 | | (if MODPDF()<0 do not use PDFLIB)| |
316 | NSTRU | Input structure function set | 5 |
317 | | (1,2=Duke-Owens1,2 3,4=EHLQ1,2 | |
318 | | 5=Owens1.1) | |
319
320 +----------+----------------------------------+-------+
321 | PRSOF | Prob. of soft underlying event | 1.0 | 07 Prsof
322
323 +----------+----------------------------------+-------+
324 | ENSOF | multipl. factor for underly evt | 1.0 | 07 Ensof
325 | PMBN1 | Soft interaction model parameter | 9.11 | 07 Pmbn1
326 | PMBN2 | Soft interaction model parameter | 0.115 | 07 Pmbn2
327 | PMBN3 | Soft interaction model parameter | -9.50 | 07 Pmbn3
328 | PMBK1 | Soft interaction model parameter | 0.029 | 07 Pmbk1
329 | PMBK2 | Soft interaction model parameter |-0.104 | 07 Pmbk2
330
331 +----------+----------------------------------+-------+
332 | PMBM1 | Soft interaction model parameter | 0.4 | 07 Pmbm1
333 | PMBM2 | Soft interaction model parameter | 2.0 | 07 Pmbm2
334 | PMBP1 | Soft interaction model parameter | 5.2 | 07 Pmbp1
335 | PMBP2 | Soft interaction model parameter | 3.0 | 07 Pmbp2
336 | PMBP3 | Soft interaction model parameter | 5.2 | 07 Pmbp3
337
338 +----------+----------------------------------+-------+
339 | IOPREM | Options for remnant clusters | 1 |
340
341 +----------+----------------------------------+-------+
342 | BTCLM | Mass param. in remnant fragmentn.| 1.0 |
343
344 +----------+----------------------------------+-------+
345 | VMIN2 | Min. parton virt-sq in dist. cal.| 0.1 |
346
347 +----------+----------------------------------+-------+
348 | CLRECO | Include colour rearrangement |.FALSE |
349 | PRECO | Probability for rearrangement | 1./9. |
350 | EXAG | Lifetime scaling for weak bosons | 1. |
351
352 +----------+----------------------------------+-------+
353 | ETAMIX | eta/eta' mixing angle in degrees | -23 |
354 | PHIMIX | phi/omega mix ang. degs | +36 |
355 | H1MIX | h1(1380)/h1(1170) mix ang. degs. | 35.26 |
356 | F0MIX | -/f0(1370) mix ang. degs. | 35.26 |
357 | F1MIX | f1(1420)/f1(1285) mix ang. degs. | 35.26 |
358 | F2MIX | f2'/f2 mix ang. degs. | +26 |
359 | ET2MIX | et2(1645)/et2(1870) mix ang. deg | 35.26 |
360 | OMHMIX | -/omega(1600) mix ang. degs. | 35.26 |
361 | PH3MIX | phi3/omega3 mix ang. degs. | +28 |
362
363 +----------+----------------------------------+-------+
364 | B1LIM | B cluster -> 1 hadron parameter | 0.0 |
365 | CLDIR(1) | Orientation of clusters, non-b | 1 | 06 Cldir1
366 | CLDIR(2) | Orientation of clusters, b | 1 | 06 Cldir2
367 | CLSMR(1) | Width of G. angle smear, non-b | 0.0 | 06 Clsmr1
368 | CLSMR(2) | Width of G. angle smear, b | 0.0 | 06 Clsmr2
369
370 +----------+----------------------------------+-------+
371 | PWT[I] | a priori weights, f.fbar pairs | 1.0 | 06 Pwt1
372 | | - I=1-7: d,u,s,c,b,t,qq' | 1.0 | 06 Pwt2
373 | | | 1.0 | 06 Pwt3
374 | | | 1.0 | 06 Pwt4
375 | | | 1.0 | 06 Pwt5
376 | | | 1.0 | 06 Pwt6
377 | | | 1.0 | 06 Pwt7
378 |REPWT(L,J,N)| a priori weight, L_J mesons | 1.0 |
379 | SNGWT | a priori wgt, singlet baryons | 1.0 | 06 Sngwt
380 | DECWT | a priori wgt, decuplet baryons | 1.0 | 06 Decwt
381
382 +----------+----------------------------------+-------+
383 | PLTCUT | Lifetime cut, "stable" particles |*1.E-11| 08 Pltcut
384
385 +----------+----------------------------------+-------+
386 | VTOCDK(I)| Veto cluster decay to hadron I |.FALSE.|
387 | VTORDK(I)| Veto resonance decay to hadron I |.FALSE.|
388 | | - I=290-293, f0(980), a0(980) | .TRUE.|
389
390 +----------+----------------------------------+-------+
391 | PIPSMR | Smear the primary vertex |.FALSE.| 01 Pipsmr
392 | VIPWID(1)| x width (mm) | 0.25 | 01 Vipwidx
393 | VIPWID(2)| y width (mm) | 0.015 | 01 Vipwidy
394 | VIPWID(3)| z width (mm) | 1.8 | 01 Vipwidz
395
396 +----------+----------------------------------+-------+
397 | MAXDKL | Veto decays outside given volume |.FALSE.|
398 | IOPDKL | Option for volume: 1=cyl, 2=sph | 1 |
399 | DXRCYL | Radius for cyl. option (mm) | 20 |
400 | DXZMAX | Length for cyl. option (mm) | 500 |
401 | DXRSPH | Radius for sph. option (mm) | 100 |
402
403 +----------+----------------------------------+-------+
404 | BDECAY | Controls which B Decay package is| 'HERW'|
405 | | used. The allowed values are: | |
406 | | 'HERW'; 'EURO'; or 'CLEO'. | |
407 | MIXING | Include neutral B meson mixing |*.FALSE.|
408 | XMIX(1) | del_Mass/Gamma, B^0_s | 10.0 |
409 | XMIX(2) | del_Mass/Gamma, B^0_d | 0.7 |
410 | YMIX(1) | del_Gamma/2*Gamma, B^0_s | 0.2 |
411 | YMIX(2) | del_Gamma/2*Gamma, B^0_d | 0.0 |
412
413 +----------+----------------------------------+-------+
414 |RMASS(198)| W+ mass | 80.42 | 11 W
415 |RMASS(199)| W- mass |W+ mass|
416 | GAMW | W+- width | 2.12 | 10 W_width
417 |RMASS(200)| Z0 mass | 91.188| 11 Z0
418 | GAMZ | Z0 width | 2.495 | 10 Z_width
419 | WZRFR | W/Z rest frame for decay pars. | .TRUE.|
420 | MODBOS(I)| Force decay modes for weak bosons| 0 | 08 Modbos1
421 | | | 0 | 08 Modbos2
422
423 +----------+----------------------------------+-------+
424 |RMASS(201)| SM Higgs mass | 115. | 11 Higgs
425 | IOPHIG | Options for large Higgs mass dist| 3 |
426 | GAMMAX | Limit on range of Higgs mass dist| 10. | 11 Gammax
427 | ENHANC(I)| Enhance factor, Higgs dk mode I | 1.0 |
428
429 +----------+----------------------------------+-------+
430 |RMASS(209)| 4th generation b quark mass | 200. |
431 |RMASS(215)| 4th generation bbar mass |RMAS209|
432
433 +----------+----------------------------------+-------+
434 | ALPHEM | Thompson limit value, alpha_em(0)|0.0072993|
435 | SWEIN | Value of sin_sq(theta_W) | 0.2319| 10 Weinberg_angle
436 | QFCH(I) | Fermion electric charge | |
437 | AFCH(I,J)| Fermion weak axial charge | |
438 | VFCH(I,J)| Fermion weak vector charge | |
439 | ZPRIME | Include a Z' in g*/Z0 processes |.FALSE.| 13 Zprime
440 |RMASS(202)| Mass of Z' | 500. | 13 ZP
441 | GAMZP | Width of Z' | 5.0 | 13 ZP_width
442 | VCKM(I,J)| CKM matrix elements | |
443 | SCABI | Value of sin_sq(theta_C) | 0.0488| 10 Cabbibo_angle
444
445 +----------+----------------------------------+-------+
446 |EPOLN(1-3)| Electron and positron beam | 0.0 |
447 | | polarizations in DIS and e+e- | 0.0 |
448 | | annihilation. First two cmpts are| 0.0 |
449 |PPOLN(1-3)| transverse and only used in e+e-,| 0.0 |
450 | | 3rd cmpt is longitudinal, and is | 0.0 |
451 | | +/-1 for fully rh/lh polarized | 0.0 |
452
453 +----------+----------------------------------+-------+
454 | QLIM | Upper limit on hard process scale| 1E08 | 04 Qlim
455
456 +----------+----------------------------------+-------+
457 | THMAX | Max thrust, IPROC=110-116 | 0.9 | 04 Thrustmax
458 | Y4JT | Min. jet separn., IPROC=600-656 | 0.01 |
459 | DURHAM | Use DURHAM or JADE, IPROC=600-656| .TRUE.|
460 |IOP4JT(1) | Color interference, IPROC=600-656| 0 |
461 |IOP4JT(2) | Color interference, IPROC=600-656| 0 |
462
463 +----------+----------------------------------+-------+
464 | BGSHAT | Scale=shat for boson-gluon fusion|.FALSE.| 04 Bgshat
465
466 +----------+----------------------------------+-------+
467 | BREIT | Use Breit frame for DIS kinematix| .TRUE.|
468 | USECMF | Use hadron-hadron cmf | .TRUE.|
469
470 +----------+----------------------------------+-------+
471 | NOSPAC | Switch off space-like showers |.FALSE.| 05 Nospac
472 | ISPAC | Changes meaning of QSPAC, | 0 | 05 Ispac
473 | | (see the earlier notes on QSPAC) | |
474
475 +----------+----------------------------------+-------+
476 | TMNISR | Min vaule shat/S for photon ISR | 1E-4 |
477 | ZMXISR | Max mom fraction for photon ISR | 1-1E-6|
478
479 +----------+----------------------------------+-------+
480 | ASFIXD | fixed alpha_s for Mueller-Tang xs| 0.25 |
481 | OMEGA | omega for Mueller-Tang xsec | 0.3 |
482
483 +----------+----------------------------------+-------+
484 | IAPHIG | Approx in Higgs+jet, IPROC=23xx | 1 |
485
486 +----------+----------------------------------+-------+
487 | PHOMAS | Damp structure functions for off-| 0.0 |
488 | | shell photons (0 for no damping) | |
489
490 +----------+----------------------------------+-------+
491 | PTMIN | Min pt in hadronic jet production| 10. | 04 Ptmin
492 | PTMAX | Max pt in hadronic jet production| 1.E8 | 04 Ptmax
493 | PTPOW | 1/pt**PTPOW for jet sampling | 4. | 04 Ptpow
494 | YJMIN | Min jet rapidity |-8. |
495 | YJMAX | Max jet rapidity | 8. | 04 Ymax
496
497 +----------+----------------------------------+-------+
498 | EMMIN | Min dilepton mass in Drell-Yan | 10. | 04 MassMin
499 | EMMAX | Max dilepton mass in Drell-Yan | 1.E8 | 04 MassMax
500 | EMPOW | 1/m**EMPOW for Drell-Yan sampling| 4. | 04 Empow
501
502 +----------+----------------------------------+-------+
503 | Q2MIN | Min Q**2 in deep inelastic | 0.0 | 04 Q2dilsmin
504 | Q2MAX | Max Q**2 in deep inelastic | 1.E10 | 04 Q2dilsmax
505 | Q2POW | (1/Q**2)**Q2POW for sampling DIS | 2.5 | 04 Q2power
506
507 +----------+----------------------------------+-------+
508 | YBMIN | Min and Max Bjorken-y in DIS | 0.0 |
509 | YBMAX | | 1.0 |
510
511 +----------+----------------------------------+-------+
512 | WHMIN |Min had mass in gamma-ind. procs. | 0.0 |
513
514 +----------+----------------------------------+-------+
515 | ZJMAX | Max Z in J/psi production | 0.9 |
516
517 +----------+----------------------------------+-------+
518 | Q2WWMN | Min Q**2 in Equiv Photon Approx | 0.0 |
519 | Q2WWMX | Max Q**2 in Equiv Photon Approx | 4.0 |
520 | YWWMIN | Min, max photon light-cone fract.| 0.0 |
521 | YWWMAX | - in Equiv. Photon approx | 1.0 |
522
523 +----------+----------------------------------+-------+
524 | CSPEED | Speed of light in vacuum (mm/s) |2.99792E11|
525 | GEV2NB | (crossh.c/e)**2 |389379.|
526
527 +----------+----------------------------------+-------+
528 | IBSH | No. of shots for init. max wgt. | 10000 |
529 | IBRN(1) | 1st random no. seed for max wgt. |1246579|
530 | IBRN(2) | - search |8447766|
531
532 +----------+----------------------------------+-------+
533 | NQEV | No. of entries in Sudakov table | 1024 |
534 | ZBINM | Max bin size, z in spacelike bran| 0.05 |
535 | NZBIN | Max no. of z bins in spacelike br| 100 |
536
537 +----------+----------------------------------+-------+
538 | NBTRY | Max. attempts to branch a parton | 200 |
539 | NCTRY | Max. attempts to decay a cluster | 200 |
540 | NETRY | Max. attempts, generate reqd mass| 200 |
541 | NSTRY | Max. attempts at soft subprocess | 200 |
542
543 +----------+----------------------------------+-------+
544 | ACCUR | Precision, soft Gauss integration| E-6 |
545
546 +----------+----------------------------------+-------+
547 | RPARTY | R-parity conservation in SUSY | .TRUE.| 12 Rparty
548
549 +----------+----------------------------------+-------+
550 | SUSYIN | Check if SUSY dataalready loaded |.FALSE.|
551 | LRSUSY | Unit for reading SUSY data | 66 |
552
553 +----------+----------------------------------+-------+
554
555
556
557 Additional quantities in HERWIG talk-to:
558
559 talk-to
560 Name Description Default Menu "name"
561
562 TLOUT time to terminate gracefully *-1e07 . 09 Tlout
563 RSTAB[21] whether pizero is stable .... false . 08 Pizstable
564
565 ..... used internally in HerwigModule.cc 0 09 No_shower
566 ..... " " " 0 09 No_hadronization
567 ..... " " " 12 File_Readsusy
568
569
570
571 Menu numbers and commands
572 G [none]
573 01 Beam
574 02 Sudakov
575 03 Struc_Function
576 04 Hards
577 05 Shower
578 06 Hadronization
579 07 Underl_event
580 08 Decays
581 09 Generation
582 10 Sm_par
583 11 Masses
584 12 Read_susy
585 13 Zprime
586 14 Prints
587
588 -------------------------------------------
589
590
591 Assorted notes.
592
593 The exact "name" is required in the talk-to. For example, Help, Show,
594 Q2pow do not work (get a brief message and the thing continues on).
595 All Boolean variables in the explicitly Herwig talk-to are there as
596 integers. So, for example, talk-to requires Bgshat set 1 ; the line
597 Bgshat set true (or, .true.) gets no error message and does not change
598 Bgshat.
599 Note that the qq talk-to uses T and F .***** may change Oct-2001 *****
600 Given a name without arguments, the program gives a simple error message
601 and continues.
602
603 For any quantity with a * in the default column above, the value given
604 is as set in HerwigInterface.F (i.e., cdf people have changed a default
605 value set by Herwig people).
606
607 Note (for anyone who worries about this) that a Herwig routine (hwuinc)
608 explicitly sets mass(W-) = mass(W+), and similarly for all six quark masses.
609 So we do not need to set both W+ and W-, both charm and anticharm masses,
610 etc.
611
612 There is a difficulty with inputting negative numbers in the talk-to.
613 For PMBN3 and PMBK2, the magnitude must be input, and the program
614 assumes the variables are negative. For example, Pmbn1 set 9.00 will lead
615 to PMBN1 = -9.00 in the HERWIG code.
616
617 The minimum jet rapidity, YJMIN above, is set to -YJMAX, the negative
618 of the maximum jet rapidity.
619
620 For inputting powers of 10, one must use E (or e), not D or d. For
621 example, Pltcut set 1.E-10 works OK, but Pltcut set 1.D-10 leads to
622 Pltcut being set to 1.
623
624 On TLOUT, 'the time to terminate gracefully'. Herwig has (in subroutine
625 hwutim, called from subr. hwuine), call timel(tres) . Then,
626 if(tres.lt.tlout) the run is ended. Subr timel is in library KERNLIB,
627 see short writeup Z007, and 'returns the execution time remaining until
628 time-limit', in seconds. By trial, the time limit (I mean the tres value
629 given by timel) starts at 999 on cdfsga for an interactive job. So, for
630 a cdfSim run taking around 12 secs per event, Herwig ends after
631 about 80 events if tlout is at the herwig default of 5.
632 I have altered the default value to -1.e07 , so presumably the condition
633 (tres.lt.tlout) will (almost) never occur. But tlout remains in the talk-to,
634 to be changed to 10/whatever if desired.
635
636 The option to write to disk a Sudakov table, and read it in, is not
637 in the talk-to. The Herwig code still allows that option, but the cpu time
638 saved is very small, so it is more sensible to recalculate the table for
639 every run.
640
641 The variable SUDORD controls whether the Sudakov form factors are
642 calculated using one-loop or two-loop alphs-s (see pages 9 and 71 of the
643 Herwig v 6 writeup). The default in Herwig is 1. In the CDF version that
644 is changed (in HerwigInterface.F) to 2, because (i) then herwig is ~40%
645 faster, (ii) it makes little difference, (iii) Mike Seymour (an author)
646 uses 2.
647
648 The variable MODBOS(I) controls the decay of the Ith gauge boson per
649 event. Some intelligence, and maybe a test run, may be needed if anyone
650 wants to change the default settings. The talkto currently allows only the
651 decays of the first two gauge bosons to be changed. The decay of W's from
652 top decay is unaffected by MODBOS settings. See the Herwig 6 writeup page
653 14 for MODBOS information.
654
655 The variable MAXEV, included in the first list above, is not used in
656 CDF's Herwig.
657
658 --------------------------------------------------------------------------
659 --------------------------------------------------------------------------
660 ***** Note Oct-2001: CHANGES have occurred and may do so again ****
661
662 Random Numbers, and Reproducing Events
663 --------------------------------------
664
665 The HERWIG package has its own random number generator, in
666 function hwrgen in hwrgen.f . This hwrgen uses two integers as seeds.
667 These seeds can be set in the Talkto - see Randseed1 and Randseed2 above.
668
669 An alternative, the CERN RANMAR random number generator, can be used
670 and exists in the CDF streams scheme. To implement that, we would
671 replace Herwig's hwrgen.f with a new hwrgen.f . Since hwrgen.f resides
672 in herwig_i/dummies the replacement would be straightforward.
673
674
675 Onto reproducibility. For example, we may want to reproduce event
676 number 3846 of a run without rerunning all 3846 events.
677 For the Herwig scheme, we just need to give the two seeds for the
678 desired event in the talk-to.
679 For RANMAR, we need to give the seeds (103 numbers) for the
680 desired event, AND we need to give the maximum weight used in the
681 earlier run. The max. weight can be set in the talk-to (Maxwt set ...),
682 and its value for the earlier run can be found in the log file, as
683 the last "NEW MAXIMUM WEIGHT" value, given with 16 decimal places. All
684 those decimal places can be used in talk-to .
685
686
687 Footnote on random numbers (for anyone who cares).
688 The CDF streams scheme is written up in CDF2459, available on the web
689 in www-cdf.fnal.gov/offline/r_n/murat/cdf_2459.html .
690 The HERWIG package generator uses a l'Ecuyer method, which combines
691 two MLCG (multiplicative linear congruential) generators. The CDF r_n
692 package uses the CERN RANMAR generator which uses a combination of a
693 fibonacci sequence and an arithmetic sequence. See the nice "A review
694 of pseudorandom number generators", F. James, Comput. Phys. Commun.
695 60, 329 (1990).
696
697 --------------------------------------------------------------------------
698 --------------------------------------------------------------------------
699
700 User routines in Herwig, and Filtering Inside Herwig
701 ----------------------------------------------------
702
703 The Herwig writeup mentions user-provided analysis routines, hwabeg,
704 hwaend, and hwanal (see page 6 of herwig v 6 write-up). Calls to hwabeg and
705 hwaend are commented out in HerwigModule.cc (in the generatorMods
706 package). There is a call to hwanal, and there is a dummy hwanal (in
707 HerwigInterface.F in the herwig_i package). hwanal (and the other two if
708 called) can provide a simple way to study generator-level quantities, and
709 to filter, i.e., select only certain events to procede with. Information is
710 all in the common block, particularly in common/hepevt/... .
711 hwanal has an integer argument, which if set to 0 will cause the
712 current event to be rejected.
713
714 A trivial hwanal.F could be
715
716 subroutine hwanal(ihwcod)
717 #include "HERWIG62.INC"
718 integer ihwcod
719 write (lhwout,10) nhep
720 10 format (' NHEP ',i8)
721 return
722 end
723
724 which will write the number of generator-level particles produced by
725 Herwig for each event.
726
727 A simple filtering hwanal.F could be
728
729 subroutine hwanal(iwhcod)
730 #include "HERWIG62.INC"
731 integer ihwcod
732 integer ii,ib
733 ib=0
734 do ii=1,nhep
735 if(idhep(ii).eq.5.or.idhep(ii).eq.-5) ib=1
736 end do
737 cc reject event unless there is at least one b quark
738 ihwcod=ib
739 return
740 end
741
742 Note that hwanal is called before any qq or tauola operation occurs.
743 So, for example, filtering on J/psi from B decays in hwanal is not possible
744 if qq is used.
745 Also, a simple parton-level filter (like the example above) would be
746 more efficient if hwanal was called before the hadronization routines
747 in HerwigModule.cc , something like the following:
748
749 // Reordered calls in HerwigModule.cc:
750 //
751 while(!goodev){
752 hwuine_(); // Initialize event
753 hwepro_(); // Generate Hard Process
754 // if(! _noshower_Parm.value()) { //skip this option for bbbar filter
755 hwbgen_(); // Generate parton cascade
756 hwdhob_(); // Do heavy quark decays
757 int ihwcod = 1;
758 hwanal_(&ihwcod);
759 if(ihwcod == 1){
760 if(! _nohadroniz_Parm.value()) {
761 hwcfor_(); // Do cluster formation
762 hwcdec_(); // Do cluster decays
763 hwdhad_(); // Do unstable particle decays
764 hwdhvy_(); // Do heavy flavor decays
765 hwmevt_(); // Add soft underlying event
766 }
767 } // end if ihwcod==1
768 // } //end if shower
769 hwufne_(); // Finish event
770 if(gHwevnt->ierror > 99) {
771 ++nterr;
772 if(nterr>mxerr) {
773 cerr << "Herwig:Too many errors on event generation" << endl;
774 goodev=true;
775 }
776 }
777 else {
778 if(ihwcod != 0) {
779 goodev=true;
780 }
781 }
782 }
783 //
784
785
786 --------------------------------------------------------------------------
787 --------------------------------------------------------------------------
788
789 Structure Function Notes
790 ------------------------
791
792 The structure functions used by Herwig are controlled by the
793 variables nstru, autpdf(i), modpdf(i), i=1,2.
794 If modpdf(i) < 0, autpdf(i) is ignored and nstru determines
795 which of the built-in structure functions is used (see hwsfun.f).
796 Allowed nstru values are 1 to 5, the default is 5, and nstru is not
797 in the talk-to.
798 If modpdf(i) >= 0, nstru is ignored, and autpdf(i) and modpdf(i)
799 determine the structure functions, which are taken from PDFLIB. Here
800 autpdf gives author initials (approx.), and modpdf gives the set number
801 (as defined by PDFLIB) for those authors. Some possible autpdf are 'MRS',
802 'CTEQ', and 'GRV', with modpdf <= 58 (MRS) or <= 41 (CTEQ) or <= 11 (GRV).
803 See the PDFLIB writeup (//consult.cern.ch/writeups/pdflib/), especially
804 Tables 1 - 4. As of August 2001 the writeup is for version 8.04 (dated
805 April 2000); one should check the current Fermilab version.
806
807 The default values in the Herwig code are modpdf(i) = -1 and autpdf(i)
808 = 'MRS'. So the default is to use nstru = 5, which is Owens set 1.1 from
809 1991, and not use PDFLIB. Because it does not seem sensible to use such an
810 old structure function set, we have changed the default for modpdf(i) to
811 41. So the default set becomes MRS set G.
812
813 The talk-to allows modpdf(i) and autpdf(i) to be changed. It is assumed
814 that i = 1 and i = 2 values are equal (OK for pbar-p). That is, talkto
815 sets Autpdf, and then autpdf(i) = Autpdf, i = 1,2, and Modpdf similarly.
816 If a non-existing (autpdf, modpdf) pair is asked for, PDFLIB defaults
817 to MRS set G. PDFLIB always writes (to log file) what set it is using.
818
819 Possible talk-to, inside talk herwig , are:
820
821 ## MRS set G:
822 Struc_Function
823 Autpdf set MRS
824 Modpdf set 41
825 exit
826 ## (this is actually the default).
827
828 ##CTEQ set 4D:
829 Struc_Function
830 Autpdf set CTEQ
831 Modpdf set 33
832 exit
833
834 --------------------------------------------------------------------------
835 --------------------------------------------------------------------------
836
837 On Decays in the Generator versus in Geant - Pltcut
838 ---------------------------------------------------
839
840 For particles with lifetimes ~ E-10 s, the question arises of whether
841 to do the decay in the generator or in Geant. If done in the generator,
842 then (i) the decay products are available in the HEPG bank, (ii) the
843 decaying particle cannot interact in the detector, (iii) if the decaying
844 particle is charged, it leaves no track and does not bend in the magnetic
845 field. If done in Geant, the converse holds.
846 Given the particle lifetimes, it seems preferable for weakly-decaying
847 strange hadrons to be decayed in geant, and B and C hadrons in the
848 generator. Therefore, pltcut, the variable used in Herwig for making
849 particles stable, has been changed (as of 26-Jan-2001) from 1.E-8 to
850 1.E-11. This change also gives agreement with the qq default of not
851 decaying K-shorts and weak-decay strange baryons. Pltcut has been added
852 to the talk-to, so if desired the previous value (or any value) can be
853 used.
854
855 --------------------------------------------------------------------------
856 --------------------------------------------------------------------------
Send problems or questions to cdfcode@fnal.gov
