Hi all,
The ROOT team is working on a new geomtry package [1], and in that
respect they called [2] for some help to test it. I put in the
subroutine GRECORD _conditionally_ on a preprocessor flags, so that
one can do the test.
Here's Rene comments on his test:
Thanks for sending the BRAHMS data. Everything OK. I have been able
to process your geometry and data points. In the attachement, you
will find a tar file with some gif files of your geometry and also
the file brahms.C automatically generated by the new g2root from
your brahms.geom file.
I have also run a test comparing the performance of our new TGeo
classes compared with Geant3 on my laptop. Here are the results;
With Geant3
===========
Time for gtmedi physics = 5.82 microseconds/call
Time for gtnext physics = 53.24 microseconds/call
Time for gtmedi random = 4.17 microseconds/call
Time for gtnext random = 29.07 microseconds/call
Time for ray tracing = 31.12 microseconds/call
with TGeo
=========
Time for FindNode physics = 5.710000 microseconds/call
Time for FindNode random = 0.520000 microseconds/call
We cannot yet run the test with distance to boundaries (end of
April). In the tests above, you should compare
gtmedi physics with FindNode physics (nearly same result)
gtmedi random with FindNode random (TGeo 8 times faster)
Your data will be very useful to us to improve the first number.
TGeo is in average twice faster than Geant3 on the LHC detectors.
By end of April, we will put the code in CVS publicly available
as well as the corresponding documentation. We are still changing
too much the code to publish anything right now.
Rene
I'll post the GIFs soon. More on that later.
The new verions of BRAT is 1.1.6.
Here's the changelog:
2002-03-21 Christian Holm Christensen <cholm@hilux15.nbi.dk>
* src/user/guxint.F, src/user/grecord.F, src/user/Makefile.am:
Added the subroutine GRECORD from Rene Brun, to help them test thier
new geometry package. It's called from GUSTEP, _conditionally_ on the
preprocessor flag TEST_ROOT_GEOM. By default, this is not defined. It
is defined if you pass the configure script the --with-root-test option.
* .cvsignore, Makefile.am: Added more ignore and clean up
* src/global/bd_mult.F: cosmetics
* configure.in: revision bump and check for ROOT test
* acinclude.m4, config/debug.m4, config/cern.m4:
Moved macroes to acinclude.m4 (easier to deal with)
* include/umedia.inc, include/umate.inc, include/ulocal.inc, include/ugeom.inc, include/udescr.inc, include/udefs.inc, include/statbit.inc, include/pawc.inc, include/m0field.inc, include/iquest.inc, include/hits.inc, include/hit_zdc.inc, include/guxfld.inc, include/gujfld.inc, include/geo_zdc.inc, include/geo_wm2.inc, include/geo_wm1.inc, include/geo_wf2.inc, include/geo_wf1.inc, include/geo_tpm2.inc, include/geo_tpm1.inc, include/geo_tofw.inc, include/geo_tof2.inc, include/geo_tof1.inc, include/geo_tile.inc, include/geo_t6.inc, include/geo_t5.inc, include/geo_t4.inc, include/geo_t3.inc, include/geo_t2.inc, include/geo_t1.inc, include/geo_supp.inc, include/geo_structure.inc, include/geo_spaa.inc, include/geo_si.inc, include/geo_rich.inc, include/geo_prea.inc, include/geo_phot.inc, include/geo_mult.inc, include/geo_mids.inc, include/geo_magnet_m0.inc, include/geo_magnet_d4.inc, include/geo_magnet_d3.inc, include/geo_magnet_d2.inc, include/geo_magnet_d1.inc, include/geo_gasc.in!
c, include/geo_fms.inc, include/geo_endc.inc, include/geo_dx.inc, include/geo_d1tr.inc, include/geo_c1.inc, include/geo_beampipe.inc, include/geo_beambeam.inc, include/gcxlun.inc, include/gcvolu.inc, include/gcuser.inc, include/gcunit.inc, include/gctrak.inc, include/gctmed.inc, include/gctime.inc, include/gcsets.inc, include/gcphys.inc, include/gconsp.inc, include/gcomis.inc, include/gcmulo.inc, include/gcmate.inc, include/gclist.inc, include/gcking.inc, include/gckine.inc, include/gcflag.inc, include/gccuts.inc, include/gcbank.inc, include/field_descrip.inc, include/evtzeb.inc, include/event.inc, include/evcode.inc, include/d4field.inc, include/d3field.inc, include/d2field.inc, include/d1field.inc:
Added mode line, and moved log to end
[1] http://root.cern.ch/root/roottalk/roottalk02/1321.html
[2] http://root.cern.ch/root/roottalk/roottalk02/1331.html
Yours,
Christian Holm Christensen -------------------------------------------
Address: Sankt Hansgade 23, 1. th. Phone: (+45) 35 35 96 91
DK-2200 Copenhagen N Cell: (+45) 28 82 16 23
Denmark Office: (+45) 353 25 305
Email: cholm@nbi.dk Web: www.nbi.dk/~cholm
This archive was generated by hypermail 2b30 : Thu Mar 21 2002 - 11:38:16 EST