Open janpipek opened 10 years ago
Hi Jan,
I don't think it is a bug, but it is something I have never figured out properly. It could be down to floating point precision/rounding and the navigator not actually being able to determine if it is on a boundary or not. It happens rarely and only in parametrised volumes it seems. At the moment I am pretending it isn't happening.
I might look into it further now that you have mentioned it.
Chris.
On 17 May 2014 23:03, Jan Pipek notifications@github.com wrote:
Dear Christopher,
this is more a question than an issue. However, I see that you have observed Geant4 calling %subj% method with negative replica number (on line 157 you deal with it) . Do you know whether this is a known bug of Geant4 (possibly round-off errors with boundaries)?
I am developing a DICOM handling library and when testing, once in a few thousand z=0 boundary crossings, this situation appears too.
Thanks, Jan
— Reply to this email directly or view it on GitHubhttps://github.com/christopherpoole/G4VoxelData/issues/15 .
Dear Christopher,
this is more a question than an issue. However, I see that you have observed Geant4 calling %subj% method with negative replica number (on line 157 you deal with it) . Do you know whether this is a known bug of Geant4 (possibly round-off errors with boundaries)?
I am developing a DICOM handling library and when testing, once in a few thousand z=0 boundary crossings, this situation appears too.
Thanks, Jan