stephenbrooks.orgForumMuon1Bug ReportsStrange determinism
Username: Password:
Search site:
Subscribe to thread via RSS
2008-01-26 14:51:33
I've been noticing some awfully similar results on the stats lists.  For example, right now, PhaseRotEb1a has four users with best results of 3.353723, and five with 3.353306. Both I and [SETI.USA] are topping Linac88MHz900MeV1 with 0.226396 while everyone else is negative (and I know I didn't inject any manual designs).  Are these truly results which have almost exactly the same number of correct-energy muons?  Or are we generating the same designs from the best-100 lists somehow?
Stephen Brooks
2008-01-27 21:00:55
The scoring counts particles, so if two similar designs let exactly the same set of particles through, they will have identical scores.
2008-01-31 07:09:13
Hmm - I was under the impression that some parts of the simulation were random (like decay time of a pion to a muon), such that this would be very unlikely.  Was it changed, or am I just imagining things? 
Stephen Brooks
2008-01-31 10:18:29
The randomness was made deterministic in version 4.44 I think.  That means for the same seed, the particles should decay in exactly the same way (it took quite a while to get this working right).
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel RSS feed

Site has had 25672266 accesses.