Hello all,
I’m a little confused by the system of batches vs particles, and the resulting calculation of standard deviation.
It’s been my understanding that for a Monte Carlo model simulating N particles, the standard deviation of the final result should be calculated from the set of all N. My apologies if I’m reading incorrectly, but I think OpenMC instead extracts the mean value for each batch and calculates the standard deviation over that set of mean values?
The OpenMC manual states “[t]he standard deviation of tally results is calculated assuming that all realizations (batches) are independent.” and this assumption of independence is mentioned elsewhere in the documentation. But what makes each batch independent of one another in the same calculation? As in, what is the difference between running 1 batch with 1000 particles, or 20 batches in the same calculation with 50 particles each? Should I expect the standard deviation of the final tally to be different in those two cases, and does that make sense?
Regards,
n.