Opened 11 years ago
Closed 11 years ago
#515 closed defect (fixed)
resurrect helicity selection rules
| Reported by: | Juergen Reuter | Owned by: | kilian |
|---|---|---|---|
| Priority: | P1 | Milestone: | v2.2.0 |
| Component: | core | Version: | 2.1.1 |
| Severity: | critical | Keywords: | |
| Cc: |
Description
WHIZARD 2.2 seems to take an order of magnitude more time to integrate: WHIZARD 2.1.1:
|=============================================================================| | WHIZARD 2.1.1 |=============================================================================| | Initializing process library 'processes' | Reading model file '/Users/reuter/local/packages/whizard-2.1.1/inst/share/whizard/models/SM.mdl' | Using model: SM | Reading commands from file 'pdf.sin' | Process 'drell_yan': ignoring existing source code | Added process to library 'processes': | [O] drell_yan = u, ubar => mu-, mu+ | Generating code for process library 'processes' | Calling O'Mega for process 'drell_yan' | command: /Users/reuter/local/packages/whizard-2.1.1/inst/bin/omega_SM.opt -o drell_yan.f90 -target:whizard -target:parameter_module parameters_SM -target:module opr_drell_yan -target:md5sum 3D4AA3CA413939FC317AA6234236BA85 -fusion:progress -scatter 'u ubar -> mu- mu+' [1/1] u ubar -> mu- mu+ ... allowed. [time: 0.00 secs, total: 0.00 secs, remaining: 0.00 secs] all processes done. [total time: 0.00 secs] SUMMARY: 6 fusions, 2 propagators, 2 diagrams | Writing interface code for process library 'processes' | Compiling process library 'processes' libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 drell_yan.f90 -fno-common -o .libs/drell_yan.o libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 drell_yan.f90 -o drell_yan.o >/dev/null 2>&1 libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 processes.f90 -fno-common -o .libs/processes.o libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 processes.f90 -o processes.o >/dev/null 2>&1 libtool: link: rm -fr .libs/processes.0.so .libs/processes.a .libs/processes.la .libs/processes.lai .libs/processes.so libtool: link: gfortran -Wl,-undefined -Wl,dynamic_lookup -o .libs/processes.0.so -bundle .libs/drell_yan.o .libs/processes.o -L/Users/reuter/local/packages/whizard-2.1.1/inst/lib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libwhizard_main.dylib -L/usr/local/lib -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0 -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0/../../.. /usr/local//lib/libHepMC.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libomega.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libomega_core.dylib -L/Users/reuter/.whizard/lib/whizard/models -L/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/models /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libwhizard.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libvamp.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libcirce1.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libcirce2.dylib /usr/local/lib/libLHAPDF.dylib /usr/local/lib/libgfortran.dylib /usr/local/lib/libquadmath.dylib -lm -lstdhep -lFmcfio -O2 libtool: link: (cd ".libs" && rm -f "processes.so" && ln -s "processes.0.so" "processes.so") libtool: link: ar cru .libs/processes.a drell_yan.o processes.o libtool: link: ranlib .libs/processes.a libtool: link: ( cd ".libs" && rm -f "processes.la" && ln -s "../processes.la" "processes.la" ) | Loading process library 'processes' | Process 'drell_yan': updating configuration sqrts = 1000.0000000000000 | Initializating integration for process drell_yan: Beam data (collision): u (mass = 0.0000000 GeV) ubar (mass = 0.0000000 GeV) sqrts = 1000.00000000000 GeV | Generating phase space configuration ... | ... done. | ... found 2 phase space channels, collected in 2 groves. | Phase space: found 2 equivalences between channels. | Wrote phase-space configuration file 'drell_yan.phs'. | iterations = 3:1000, 3:10000 Warning: No cuts have been defined. | Creating VAMP integration grids: | Using phase-space channel equivalences. | 1000 calls, 2 channels, 2 dimensions, 20 bins, stratified = T | Integrating process 'drell_yan': |=============================================================================| | It Calls Integral[fb] Error[fb] Err[%] Acc Eff[%] Chi2 N[It] | |=============================================================================| 1 1000 1.9320282E+01 3.78E-02 0.20 0.06* 36.88 2 1000 1.9299581E+01 2.09E-02 0.11 0.03* 68.24 3 1000 1.9282715E+01 2.72E-02 0.14 0.04 52.54 |-----------------------------------------------------------------------------| 3 3000 1.9297655E+01 1.52E-02 0.08 0.04 52.54 0.33 3 |-----------------------------------------------------------------------------| | Process drell_yan: Using integration grids from iteration #2 | Reading integration grids and results from file 'drell_yan.vgb': 4 10000 1.9306017E+01 2.83E-03 0.01 0.01* 68.22 5 10000 1.9304344E+01 2.00E-03 0.01 0.01* 48.26 6 10000 1.9302682E+01 1.61E-03 0.01 0.01* 60.60 |-----------------------------------------------------------------------------| 6 30000 1.9303774E+01 1.14E-03 0.01 0.01 60.60 0.59 3 |-----------------------------------------------------------------------------| |=============================================================================| 6 30000 1.9303774E+01 1.14E-03 0.01 0.01 60.60 0.59 3 |=============================================================================| | Process 'drell_yan': | time estimate for generating 10000 unweighted events: 0s |-----------------------------------------------------------------------------| | There were no errors and 1 warning(s). | WHIZARD run finished. |=============================================================================| real 0m2.135s user 0m1.668s sys 0m0.303s
vs. WHIZARD 2.2.0:
|=============================================================================| | WHIZARD 2.2.0 |=============================================================================| Warning: Rebuild options not re-implemented yet | Reading model file '/Users/reuter/local/share/whizard/models/SM.mdl' | Using model: SM | Preloaded library: default_lib | Reading commands from file 'pdf_builtin.sin' make: Nothing to be done for `source'. /Users/reuter/local/lib/whizard/libtool --tag=FC --mode=compile gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/lib/whizard/mod/models -I/Users/reuter/local/include/omega -I/Users/reuter/local/lib/whizard/mod/whizard-core -I/Users/reuter/local/lib/whizard/mod/misc -fopenmp -g -O2 -fno-common default_lib.f90 libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/lib/whizard/mod/models -I/Users/reuter/local/include/omega -I/Users/reuter/local/lib/whizard/mod/whizard-core -I/Users/reuter/local/lib/whizard/mod/misc -fopenmp -g -O2 -fno-common default_lib.f90 -fno-common -o .libs/default_lib.o libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/lib/whizard/mod/models -I/Users/reuter/local/include/omega -I/Users/reuter/local/lib/whizard/mod/whizard-core -I/Users/reuter/local/lib/whizard/mod/misc -fopenmp -g -O2 -fno-common default_lib.f90 -o default_lib.o >/dev/null 2>&1 /Users/reuter/local/lib/whizard/libtool --tag=FC --mode=link gfortran -module -rpath /dev/null -fopenmp -g -O2 -L/Users/reuter/local/lib -lwhizard_main -lwhizard -lomega -lHepMC -o default_lib.la default_lib.lo drell_yan_i1.lo libtool: link: rm -fr .libs/default_lib.0.so .libs/default_lib.a .libs/default_lib.la .libs/default_lib.lai .libs/default_lib.so libtool: link: gfortran -Wl,-undefined -Wl,dynamic_lookup -o .libs/default_lib.0.so -bundle .libs/default_lib.o .libs/drell_yan_i1.o -L/Users/reuter/local/lib /Users/reuter/local/lib/libwhizard_main.dylib /Users/reuter/local/lib/libwhizard.dylib -L/usr/local/lib -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0 -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0/../../.. /Users/reuter/local/lib/libvamp.dylib /Users/reuter/local/lib/libcirce1.dylib /Users/reuter/local/lib/libcirce2.dylib /usr/local/lib/libLHAPDF.dylib /usr/local/lib/libgfortran.dylib /usr/local/lib/libquadmath.dylib -lm -lstdhep -lFmcfio /Users/reuter/local/lib/libomega.dylib /Users/reuter/local/lib/libomega_core.dylib /usr/local//lib/libHepMC.dylib -fopenmp -O2 -fopenmp libtool: link: (cd ".libs" && rm -f "default_lib.so" && ln -s "default_lib.0.so" "default_lib.so") libtool: link: ar cru .libs/default_lib.a default_lib.o drell_yan_i1.o libtool: link: ranlib .libs/default_lib.a libtool: link: ( cd ".libs" && rm -f "default_lib.la" && ln -s "../default_lib.la" "default_lib.la" ) sqrts = 1.000000000000E+03 Beam data (collision): u (mass = 0.0000000E+00 GeV) ubar (mass = 0.0000000E+00 GeV) sqrts = 1.000000000000E+03 GeV | Beams: u ubar | Beams: sqrts = 1.000E+03 GeV | Phase space: generating configuration ... | Phase space: ... success. | Phase space: writing configuration file 'drell_yan_i1.phs' | Phase space: 2 channels, 2 dimensions | Phase space: wood | Starting integration for process 'drell_yan' | Integrate: iterations not specified, using default | Integrator: 2 chains, 2 channels, 2 dimensions | Integrator: VAMP |=============================================================================| | It Calls Integral[fb] Error[fb] Err[%] Acc Eff[%] Chi2 N[It] | |=============================================================================| 1 784 1.9309897E+01 3.90E-02 0.20 0.06* 0.00 2 784 1.9308256E+01 2.79E-02 0.14 0.04* 0.00 3 784 1.9343524E+01 2.81E-02 0.15 0.04 0.00 |-----------------------------------------------------------------------------| 3 2352 1.9320559E+01 1.99E-02 0.10 0.05 0.00 0.00 3 |-----------------------------------------------------------------------------| 4 9936 1.9303810E+01 3.00E-03 0.02 0.02* 0.00 5 9936 1.9300514E+01 2.99E-03 0.02 0.02* 0.00 6 9936 1.9303700E+01 3.03E-03 0.02 0.02 0.00 |-----------------------------------------------------------------------------| 6 29808 1.9302674E+01 1.87E-03 0.01 0.02 0.00 0.00 3 |=============================================================================| | There were no errors and 1 warning(s). | WHIZARD run finished. |=============================================================================| real 0m12.302s user 0m4.639s sys 0m12.595s
Especially the last three iterations are really slow!!!
Change History (11)
comment:1 Changed 11 years ago by
| Priority: | P0 → P1 |
|---|---|
| Status: | new → assigned |
| Summary: | Integration profiling → resurrect helicity selection rules |
comment:2 Changed 11 years ago by
Touche!!! 2.1.1:
| Phase space: generating configuration ... | Phase space: ... failed. Increasing phs_off_shell ... | Phase space: ... success. | Phase space: writing configuration file 'platt_i1.phs' | Phase space: 27 channels, 5 dimensions | Phase space: wood | Starting integration for process 'platt' | Integrate: iterations not specified, using default | Integrator: 3 chains, 27 channels, 5 dimensions | Integrator: VAMP |=============================================================================| | It Calls Integral[fb] Error[fb] Err[%] Acc Eff[%] Chi2 N[It] | |=============================================================================| 1 4995 1.0983564E-03 2.88E-06 0.26 0.19* 0.00 2 4983 1.1007610E-03 2.01E-06 0.18 0.13* 0.00 3 4977 1.0978265E-03 1.59E-06 0.14 0.10* 0.00 4 4968 1.0979320E-03 1.44E-06 0.13 0.09* 0.00 5 4956 1.0982908E-03 1.35E-06 0.12 0.09* 0.00 |-----------------------------------------------------------------------------| 5 24879 1.0986333E-03 1.21E-06 0.11 0.17 0.00 0.00 5 |-----------------------------------------------------------------------------| 6 9990 1.0982890E-03 9.81E-07 0.09 0.09 0.00 7 9990 1.0981485E-03 9.56E-07 0.09 0.09* 0.00 8 9990 1.0978401E-03 1.00E-06 0.09 0.09 0.00 |-----------------------------------------------------------------------------| 8 29970 1.0980925E-03 2.30E-07 0.02 0.04 0.00 0.00 3 |=============================================================================| | There were no errors and 1 warning(s). | WHIZARD run finished. |=============================================================================| real 0m22.319s user 0m11.436s sys 0m23.917s
2.2.0:
| Phase space: generating configuration ... | Phase space: ... failed. Increasing phs_off_shell ... | Phase space: ... success. | Phase space: writing configuration file 'platt_i1.phs' | Phase space: 27 channels, 5 dimensions | Phase space: wood | Starting integration for process 'platt' | Integrate: iterations not specified, using default | Integrator: 3 chains, 27 channels, 5 dimensions | Integrator: VAMP |=============================================================================| | It Calls Integral[fb] Error[fb] Err[%] Acc Eff[%] Chi2 N[It] | |=============================================================================| 1 4995 1.0983717E-03 2.88E-06 0.26 0.19* 0.00 2 4983 1.1006926E-03 2.01E-06 0.18 0.13* 0.00 3 4977 1.0978768E-03 1.59E-06 0.14 0.10* 0.00 4 4968 1.0980335E-03 1.44E-06 0.13 0.09* 0.00 5 4956 1.0982987E-03 1.36E-06 0.12 0.09* 0.00 |-----------------------------------------------------------------------------| 5 24879 1.0986546E-03 1.16E-06 0.11 0.17 0.00 0.00 5 |-----------------------------------------------------------------------------| 6 9993 1.0981799E-03 9.80E-07 0.09 0.09 0.00 7 9993 1.0981250E-03 9.58E-07 0.09 0.09* 0.00 8 9993 1.0977251E-03 1.00E-06 0.09 0.09 0.00 |-----------------------------------------------------------------------------| 8 29979 1.0980100E-03 2.48E-07 0.02 0.04 0.00 0.00 3 |=============================================================================| | There were no errors and 1 warning(s). | WHIZARD run finished. |=============================================================================| real 0m22.041s user 0m11.322s sys 0m23.685s
(this was for bb -> 3H with mb = 10 GeV, identical SINDARIN input files)
comment:4 Changed 11 years ago by
Nope, here is the result:
|=============================================================================| | WHIZARD 2.1.1 |=============================================================================| | Initializing process library 'processes' | Loading process library 'processes' | [No compiled library 'processes'] | Reading model file '/Users/reuter/local/packages/whizard-2.1.1/inst/share/whizard/models/SM.mdl' | Using model: SM | Reading commands from file 'pdf.sin' | Added process to library 'processes': | [O] platt = b, bbar => H, H, H | Generating code for process library 'processes' | Calling O'Mega for process 'platt' | command: /Users/reuter/local/packages/whizard-2.1.1/inst/bin/omega_SM.opt -o platt.f90 -target:whizard -target:parameter_module parameters_SM -target:module opr_platt -target:md5sum FDF0BB8D53879F2433E3DA3B7DD7CC3A -fusion:progress -scatter 'b bbar -> H H H' [1/1] b bbar -> H H H ... allowed. [time: 0.00 secs, total: 0.00 secs, remaining: 0.00 secs] all processes done. [total time: 0.00 secs] SUMMARY: 34 fusions, 10 propagators, 16 diagrams | Writing interface code for process library 'processes' | Compiling process library 'processes' libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 platt.f90 -fno-common -o .libs/platt.o libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 platt.f90 -o platt.o >/dev/null 2>&1 libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 processes.f90 -fno-common -o .libs/processes.o libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/models -I/Users/reuter/local/packages/whizard-2.1.1/inst/include/omega -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/whizard-core -I/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/mod/misc -g -O2 processes.f90 -o processes.o >/dev/null 2>&1 libtool: link: gfortran -Wl,-undefined -Wl,dynamic_lookup -o .libs/processes.0.so -bundle .libs/platt.o .libs/processes.o -L/Users/reuter/local/packages/whizard-2.1.1/inst/lib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libwhizard_main.dylib -L/usr/local/lib -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0 -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0/../../.. /usr/local//lib/libHepMC.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libomega.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libomega_core.dylib -L/Users/reuter/.whizard/lib/whizard/models -L/Users/reuter/local/packages/whizard-2.1.1/inst/lib/whizard/models /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libwhizard.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libvamp.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libcirce1.dylib /Users/reuter/local/packages/whizard-2.1.1/inst/lib/libcirce2.dylib /usr/local/lib/libLHAPDF.dylib /usr/local/lib/libgfortran.dylib /usr/local/lib/libquadmath.dylib -lm -lstdhep -lFmcfio -O2 libtool: link: (cd ".libs" && rm -f "processes.so" && ln -s "processes.0.so" "processes.so") libtool: link: ar cru .libs/processes.a platt.o processes.o libtool: link: ranlib .libs/processes.a libtool: link: ( cd ".libs" && rm -f "processes.la" && ln -s "../processes.la" "processes.la" ) | Loading process library 'processes' | Process 'platt': updating configuration sqrts = 1000.0000000000000 SM.mb = 10.000000000000000 | Initializating integration for process platt: Beam data (collision): b (mass = 10.000000 GeV) bbar (mass = 10.000000 GeV) sqrts = 1000.00000000000 GeV | Phase space file 'platt.phs' not found. | Generating phase space configuration ... | ... done. | ... found 27 phase space channels, collected in 3 groves. | Phase space: found 234 equivalences between channels. | Wrote phase-space configuration file 'platt.phs'. | iterations = 5:5000, 3:10000 Warning: No cuts have been defined. | Creating VAMP integration grids: | Using phase-space channel equivalences. | 5000 calls, 27 channels, 5 dimensions, 18 bins, stratified = T | Integrating process 'platt': |=============================================================================| | It Calls Integral[fb] Error[fb] Err[%] Acc Eff[%] Chi2 N[It] | |=============================================================================| 1 5000 1.0942904E-03 2.91E-06 0.27 0.19* 61.63 2 5000 1.0966324E-03 1.72E-06 0.16 0.11* 68.75 3 5000 1.1005713E-03 1.69E-06 0.15 0.11* 72.36 4 5000 1.0995929E-03 1.51E-06 0.14 0.10* 75.71 5 5000 1.0983297E-03 1.45E-06 0.13 0.09* 78.56 |-----------------------------------------------------------------------------| 5 25000 1.0984954E-03 7.63E-07 0.07 0.11 78.56 1.33 5 |-----------------------------------------------------------------------------| 6 10000 1.0983119E-03 1.05E-06 0.10 0.10 73.30 7 10000 1.0990859E-03 6.97E-07 0.06 0.06* 69.91 8 10000 1.0985494E-03 5.64E-07 0.05 0.05* 76.18 |-----------------------------------------------------------------------------| 8 30000 1.0986945E-03 4.04E-07 0.04 0.06 76.18 0.26 3 |-----------------------------------------------------------------------------| |=============================================================================| 8 30000 1.0986945E-03 4.04E-07 0.04 0.06 76.18 0.26 3 |=============================================================================| | Process 'platt': | time estimate for generating 10000 unweighted events: 0s |-----------------------------------------------------------------------------| | There were no errors and 1 warning(s). | WHIZARD run finished. |=============================================================================| real 0m4.228s user 0m3.749s sys 0m0.308s
and
|=============================================================================| | WHIZARD 2.2.0 |=============================================================================| Warning: Rebuild options not re-implemented yet | Reading model file '/Users/reuter/local/share/whizard/models/SM.mdl' | Using model: SM | Preloaded library: default_lib | Reading commands from file 'pdf.sin' make: Nothing to be done for `source'. /Users/reuter/local/lib/whizard/libtool --tag=FC --mode=compile gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/lib/whizard/mod/models -I/Users/reuter/local/include/omega -I/Users/reuter/local/lib/whizard/mod/whizard-core -I/Users/reuter/local/lib/whizard/mod/misc -fopenmp -g -O2 -fno-common default_lib.f90 libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/lib/whizard/mod/models -I/Users/reuter/local/include/omega -I/Users/reuter/local/lib/whizard/mod/whizard-core -I/Users/reuter/local/lib/whizard/mod/misc -fopenmp -g -O2 -fno-common default_lib.f90 -fno-common -o .libs/default_lib.o libtool: compile: gfortran -c -I/Users/reuter/.whizard/lib/whizard/mod/models -I/Users/reuter/local/lib/whizard/mod/models -I/Users/reuter/local/include/omega -I/Users/reuter/local/lib/whizard/mod/whizard-core -I/Users/reuter/local/lib/whizard/mod/misc -fopenmp -g -O2 -fno-common default_lib.f90 -o default_lib.o >/dev/null 2>&1 /Users/reuter/local/lib/whizard/libtool --tag=FC --mode=link gfortran -module -rpath /dev/null -fopenmp -g -O2 -L/Users/reuter/local/lib -lwhizard_main -lwhizard -lomega -lHepMC -o default_lib.la default_lib.lo platt_i1.lo libtool: link: rm -fr .libs/default_lib.0.so .libs/default_lib.a .libs/default_lib.la .libs/default_lib.lai .libs/default_lib.so libtool: link: gfortran -Wl,-undefined -Wl,dynamic_lookup -o .libs/default_lib.0.so -bundle .libs/default_lib.o .libs/platt_i1.o -L/Users/reuter/local/lib /Users/reuter/local/lib/libwhizard_main.dylib /Users/reuter/local/lib/libwhizard.dylib -L/usr/local/lib -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0 -L/usr/local/lib/gcc/x86_64-apple-darwin11.4.2/4.9.0/../../.. /Users/reuter/local/lib/libvamp.dylib /Users/reuter/local/lib/libcirce1.dylib /Users/reuter/local/lib/libcirce2.dylib /usr/local/lib/libLHAPDF.dylib /usr/local/lib/libgfortran.dylib /usr/local/lib/libquadmath.dylib -lm -lstdhep -lFmcfio /Users/reuter/local/lib/libomega.dylib /Users/reuter/local/lib/libomega_core.dylib /usr/local//lib/libHepMC.dylib -fopenmp -O2 -fopenmp libtool: link: (cd ".libs" && rm -f "default_lib.so" && ln -s "default_lib.0.so" "default_lib.so") libtool: link: ar cru .libs/default_lib.a default_lib.o platt_i1.o libtool: link: ranlib .libs/default_lib.a libtool: link: ( cd ".libs" && rm -f "default_lib.la" && ln -s "../default_lib.la" "default_lib.la" ) sqrts = 1.000000000000E+03 SM.mb = 1.000000000000E+01 Beam data (collision): b (mass = 1.0000000E+01 GeV) bbar (mass = 1.0000000E+01 GeV) sqrts = 1.000000000000E+03 GeV | Beams: b bbar | Beams: sqrts = 1.000E+03 GeV | Phase space: generating configuration ... | Phase space: ... failed. Increasing phs_off_shell ... | Phase space: ... success. | Phase space: writing configuration file 'platt_i1.phs' | Phase space: 27 channels, 5 dimensions | Phase space: wood | Starting integration for process 'platt' | Integrate: iterations not specified, using default | Integrator: 3 chains, 27 channels, 5 dimensions | Integrator: VAMP |=============================================================================| | It Calls Integral[fb] Error[fb] Err[%] Acc Eff[%] Chi2 N[It] | |=============================================================================| 1 4995 1.0984144E-03 2.89E-06 0.26 0.19* 0.00 2 4983 1.1007007E-03 2.01E-06 0.18 0.13* 0.00 3 4977 1.0979692E-03 1.59E-06 0.14 0.10* 0.00 4 4968 1.0980030E-03 1.44E-06 0.13 0.09* 0.00 5 4956 1.0982669E-03 1.36E-06 0.12 0.09* 0.00 |-----------------------------------------------------------------------------| 5 24879 1.0986708E-03 1.15E-06 0.10 0.17 0.00 0.00 5 |-----------------------------------------------------------------------------| 6 9993 1.0981616E-03 9.80E-07 0.09 0.09 0.00 7 9993 1.0980295E-03 9.58E-07 0.09 0.09* 0.00 8 9993 1.0977460E-03 1.00E-06 0.09 0.09 0.00 |-----------------------------------------------------------------------------| 8 29979 1.0979790E-03 2.12E-07 0.02 0.03 0.00 0.00 3 |=============================================================================| | There were no errors and 1 warning(s). | WHIZARD run finished. |=============================================================================| real 0m21.976s user 0m11.283s sys 0m23.483s
BTW, why these non-round numbers for the number of calls?
comment:5 Changed 11 years ago by
OK, I'll check anyway. So let's compile profiled versions.
BTW: OpenMP was turned off for both versions, right?
comment:6 Changed 11 years ago by
WHIZARD 2.1.1 was compiled w\ OpenMP, WHIZARD 2.2.0 w/ OpenMP. But I guess it's not functional in 2.2.0 anyhow. So there still remains a factor of 5.
comment:7 Changed 11 years ago by
I got some data points, too:
H, H => H, H, H (5x5000 + 3x300,000): 30.4s (2.1.1) vs. 36.2s (2.2.0) b, B => H, H, H (5x5000 + 3x300,000): 36.8s (2.1.1) vs. 43.6s (2.2.0) u, U => e2, E2 (5x5000 + 3x300,000): 13.6s (2.1.1) vs. 12.8s (2.2.0)
I observe only a small difference. There is some room for improvement in the first two processes, but no need to worry.
Both setups have OpenMP switched off and helicity selection switched off. Compiler is gfortran -g -O2, version 4.6.3 The timings include process generation, compilation, phase space setup and integration, but the 300k integration blocks dominate.
Could you double-check?
comment:8 Changed 11 years ago by
My numbers with gfortran 4.9.0 (experimental) -g -O2 OpenMP switched off are:
H, H => H, H, H (5x5000 + 3x300,000): 29.1s (2.1.1) vs. 33.8s (2.2.0) b, B => H, H, H (5x5000 + 3x300,000): 37.4s (2.1.1) vs. 40.7s (2.2.0) u, U => e2, E2 (5x5000 + 3x300,000): 13.3s (2.1.1) vs. 9.8s (2.2.0)
Version 0, edited 11 years ago
by
(next)
comment:9 Changed 11 years ago by
| Severity: | blocker → critical |
|---|
So the current state is that helicity selection should be resurrected (I think I can do it soon), but otherwise it is a matter of fine-tuning, not an order-of-magnitude effect.
BTW, for proper profiling I need static executables, which I deliberately postponed. Let's discuss this in the next meeting.
comment:11 Changed 11 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | assigned → closed |
Yes, this was presumably done in r4564.
Note: See
TracTickets for help on using
tickets.
Not unexpected. If I'm not completely off track, the helicity selection counting of O'Mega has not been resurrected yet. I'd expect a factor 4 for this process, right? Maybe you could check a process without selection rules, e.g., b bbar (massive) => 3H?
(I assume that only the 'user' part is relevant, the 'sys' part shouldn't be attributed to the program.)
But the ticket is important as a reminder, I keep it but change the title.