On Tue, Sep 01, 2015 at 11:25:34PM +1000, Rick Walsh wrote: > > I see perhaps the greatest improvement is that it should make calculating a > decent VPM-B ceiling for a real dive one step closer.
That would be neat :-) > > This results in slightly different runtimes in the test dives (typically a > > minute compared to what we did previously). The last patch updates the test > > but compared to the FORTRAN code, we now fail two tests. So this is up to > > discussion. > > It looks like Subsurface is now more "accurate" and consistent than the > Fortran code, which is nice. But it also looks like Subsurface is > marginally more aggressive in most cases, which is not so good. The > Fortran code, and it's more modern but seemingly barely changed > derivations, is and for the near future will probably stay the standard > reference, and arguably the "model" incorporates this when used for real > dives. If you spend an extra minute or two at 6 m, your body won't care > what model you used to plan your dive. Let's please remember that these algorithms are mathematical approximations of mostly unproven theories with random, made up limits that were tuned to make them match known good profiles. And then a safety factor was added for good measurement. So "accurate" is a very ill defined term here. That said, it would be a nice comfort booster if we did match the numbers by the reference algorithm exactly or at least didn't show any shorter deco (simply from a "risk of some idiot suing us" perspective. > Assuming there's no mistake that needs fixing in our implementation, we > could either live with the difference, or we could introduce a slight > conservatism to our nominal case. The following is a very basic patch that > makes us pass all the tests again. I'm not proposing it yes, at least not > until we've worked out there isn't some other tweak that we ought to do. > But it does work. Um. On some level this makes me cringe (no offense Rick - this has nothing to do with your code) - it's yet one more "random number pulled out of a dark place to make profiles match. But given what I said above, I'd be certainly willing to take this simply as insurance that we don't end up with shorter deco times than the reference. Robert, what do you think? /D _______________________________________________ subsurface mailing list [email protected] http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
