5 ms·
I think that you're smart enough to grok a slight nuance to point 1; it's a Monte Carlo simulation. Being non-repeatable and having to average results is part o
by lidHanteyk 6y ago
I think that you're smart enough to grok a slight nuance to point 1; it's a Monte Carlo simulation. Being non-repeatable and having to average results is part of the definition of the simulation.
All code is trash, but not all code is trash for all the same reasons.
- WalterBright 6y agoThe reason for its random behavior is not Monte Carlo, it's because of bugs in the program. Please read the article.
- selimthegrim 6y agoIf you read further you would have seen that the bug was of limited duration (saved state from a PRNG), the model reproduced upon a second run (third from the start) and furthermore didn't effect the results since this is ensemble modeling. As a fellow Techer I expected that you would have read all the responses before passing judgement.
- thu2111 6y agoThe code isn't a Monte Carlo simulation if you read it. It just iterates until it's simulated every day of the epidemic with the input seeds and then outputs the results to a table. The results are then given exactly without any uncertainty bounds: to make it a Monte Carlo simulation is probably impossible given how slow it is. I think the article author mentioned that just to show (s)he understands stochastic modelling. As can be seen from your comments and others, it's apparent some people/scientists either aren't reading the article properly at all, or have a massively wrong understanding of what Monte Carlo/stochastic modelling techniques really are.