[Gate-users] Can't run with 100 million particles

Thu Oct 23 17:13:28 CEST 2014

Hi Amy,

I'm not sure, but I think this paper is one of the references mentioned:

I.J. Chetty et al., “Reporting and analyzing statistical uncertainties in Monte Carlo-based treatment planning.,” Int. J. Radiat. Oncol. Biol. Phys. 65(4), 1249–1259 (2006).

You may want to look at this paper for the history-by-history statistical uncertainty estimation method (referenced in the previous paper):

B.R.B. Walters, I. Kawrakow, and D.W.O. Rogers, “History by history statistical estimators in the BEAM code system,” Med. Phys. 29(12), 2745 (2002).

It might be a bit tricky to wrap your head around that method, especially with all the Monte Carlo terminology thrown around (batch, history, recycling, etc.)

Marc

__________________________

Marc Chamberland, MSc
PhD candidate
Department of Physics
Carleton University
Ottawa (ON)

On Oct 23, 2014, at 09:17, Amy Meldrum <ameldru at g.clemson.edu<mailto:ameldru at g.clemson.edu>> wrote:

Can anyone point me towards any of the reference papers that were mentioned in this thread? I think they will be useful for me in the near future when I defend my Master's thesis.

Thanks,

Amy

On Wed, Oct 15, 2014 at 6:40 AM, Grevillot Loic <loic.grevillot at medaustron.at<mailto:loic.grevillot at medaustron.at>> wrote:
Of course not!
If you buy 2 times 10 apples or 1 time 20 apples you get in both cases 20 apples.
Reference to papers used for uncertainty calculations are available – please look at Gate reference papers!
Please use the list, other people may benefit of discussion and my contributions are on free time which is very limited at the moment…

Kind regards,

Loïc Grevillot

From: S. Saghamanesh [mailto:s.saghamanesh at yahoo.com<mailto:s.saghamanesh at yahoo.com>]
Sent: Mittwoch, 15. Oktober 2014 12:34
To: Grevillot Loic
Subject: Re: [Gate-users] Can't run with 100 million particles

Thanks for your note. However, due to the linear operation of this addition, I'll have a greater uncertainty!
For example, I have two simulations with uncertainties of sigma1 = 0.7 (70%) and sigma2 = 0.6 (60%). After the addition, I'll have the total uncertainty of sqrt(sigma1^2 + sigma2^2) = 0.92 (92%).
So I wanted to reduce my uncertainty by increasing my statistics (through several simulations), but ultimately I'd obtain a greater uncertainty. Am I right on this?

Best
Soman

On Monday, October 13, 2014 8:41 AM, Grevillot Loic <loic.grevillot at medaustron.at<mailto:loic.grevillot at medaustron.at>> wrote:

You can manually run 10 different simulations. Make sure to change output names for each simu and then make a simple addition.

Sincerely,

Loïc

From: S. Saghamanesh [mailto:s.saghamanesh at yahoo.com]
Sent: Samstag, 11. Oktober 2014 13:53
To: Grevillot Loic; gate-users at lists.opengatecollaboration.org<mailto:gate-users at lists.opengatecollaboration.org>
Subject: Re: [Gate-users] Can't run with 100 million particles

Dear Loïc,

I've tried to run my main macro file through the following command:
 "/control/loop myMain.mac mycounter 1 10 10"
which is supposed to run 10 simulations with 10 million particles with different seeds (by /gate/random/setEngineSeed auto). But It runs only the first 10 million particles and doesn't keep run the remains.
Do you have any idea what's wrong with this command? I guess it's the only way I can proceed with 100 million primaries without getting error.

Best regards
Soman

On Friday, October 3, 2014 9:10 AM, Grevillot Loic <loic.grevillot at medaustron.at<mailto:loic.grevillot at medaustron.at>> wrote:

Dear Soman,

You said you need to run 100 million primaries and artifacts appear after 10 million.
My suggestion is therefore to run 10 simulations of 10 million particles (fully equivalent to a single simu of 100 million). You just have to make sure the seeds are different for every simulation, which can be easily achieved using the following command:
/gate/random/setEngineSeed auto

Best,
Loïc

From: S. Saghamanesh [mailto:s.saghamanesh at yahoo.com]
Sent: Freitag, 3. Oktober 2014 07:30
To: Grevillot Loic
Subject: Re: [Gate-users] Can't run with 100 million particles

Hi,

Thanks a lot for your note. I use of splitting 1000 for my emitted photons but I still have unacceptable uncertainties, so I have to run with 100 million particles. Do you tell me if I run 10 simulations with 10 million particles, it would lessen the uncertainty? I guess it wouldn't, because I have uncertain results with a single simulation of 10 million particles.

Thanks in advance for your help
- Soman

On Wednesday, October 1, 2014 3:41 PM, Grevillot Loic <loic.grevillot at medaustron.at<mailto:loic.grevillot at medaustron.at>> wrote:

Dear Soman,

We noticed a similar simulation “artifact” when a too high number of particles was used. We also noticed that the effect is dependent on the particle type and simulation parameters (cut, step size).
We will try correcting the bug for the next release. In the meantime I advise you to run 10 simulations of 10 million particles.

Sincerely,

Loïc

From: gate-users-bounces at lists.opengatecollaboration.org<mailto:gate-users-bounces at lists.opengatecollaboration.org> [mailto:gate-users-bounces at lists.opengatecollaboration.org] On Behalf Of S. Saghamanesh
Sent: Mittwoch, 1. Oktober 2014 09:29
To: gate-users at lists.opengatecollaboration.org<mailto:gate-users at lists.opengatecollaboration.org>
Subject: [Gate-users] Can't run with 100 million particles

Hi,

Is there any limitation on the number of primary particles generated in Gate7.0 ? It seems running code with more than 10 million particles leads to false results. I need to run my code with 100 millions particles, but after checking (with a very simple example), I noticed that no more than 20 million particles would be generated.

Thanks for your comments.
- Soman
Disclaimer:
Please notice our E-Mail Disclaimer http://www.medaustron.at/email-disclaimer/

Disclaimer:
Please notice our E-Mail Disclaimer http://www.medaustron.at/email-disclaimer/

Disclaimer:
Please notice our E-Mail Disclaimer http://www.medaustron.at/email-disclaimer/

Disclaimer:
Please notice our E-Mail Disclaimer http://www.medaustron.at/email-disclaimer/

_______________________________________________
Gate-users mailing list
Gate-users at lists.opengatecollaboration.org<mailto:Gate-users at lists.opengatecollaboration.org>
http://lists.opengatecollaboration.org/mailman/listinfo/gate-users

_______________________________________________
Gate-users mailing list
Gate-users at lists.opengatecollaboration.org<mailto:Gate-users at lists.opengatecollaboration.org>
http://lists.opengatecollaboration.org/mailman/listinfo/gate-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/mailman/private/gate-users/attachments/20141023/bf61e4f0/attachment-0001.html>