Problem 605 - extremely slow Penelope Rayleigh computation
Summary: extremely slow Penelope Rayleigh computation
Status: RESOLVED FIXED
Alias: None
Product: Geant4
Classification: Unclassified
Component: processes/electromagnetic/lowenergy (show other problems)
Version: 6.0
Hardware: Other Other
: P2 normal
Assignee: Luciano Pandola
URL:
Depends on:
Blocks:
 
Reported: 2004-03-17 13:13 CET by marcus.h.mendenhall
Modified: 2004-03-18 07:10 CET (History)
0 users

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this problem.
Description marcus.h.mendenhall 2004-03-17 13:13:33 CET 
In G4PenelopeRayleigh, the re-computation of the random sampling function for each
call to PostStepDoIt results in my entire simulation being slowed down a factor of more
than 30.  I am attaching a modification to G4PenelopeRayleigh.cc and the associated
header file which caches these results, and looks up re-used values.  This eleminates
the bottleneck.

These should be verified for correctness, and if OK, (and someone else hasn't already
fixed this internally), checked in.

Thanks.
Comment 1 Maria.Grazia.Pia 2004-03-18 03:17:59 CET 
The problem is re-assigned to the developer expert in Geant4 Penelope processes.

Best wishes,
Maria Grazia Pia
Comment 2 Luciano Pandola 2004-03-18 07:10:59 CET 
Your solution is correct indeed and it can greatly reduce the computation time
for the Rayleigh process. The modified process will be available in the next
public release/patch. Thank you very much for your report and solution.
Best wishes,
Luciano Pandola