RE: Question about the deviation

Hi Lee,

Andrea is right, FLUKA reports the statistical error in percent in the ASCII output files.
So your absolute error is: value * error / 100.0
Therefore 4.9309e-3% is EXTREMELY small, which means your are dominated by
the systematic uncertainty of the simulation and not the statistical one.
Thus no need to run so many cycles!

Please look in the course lecture on statistics and sampling (especially page 24)
https://www.fluka.org/free_download/course/prague2011/Lectures/MCstat1211.pdf

Vasilis
Received on Thu Apr 05 2012 - 17:37:07 CEST