On Thu, Mar 3, 2016 at 1:17 PM, Lubomir I. Ivanov <[email protected]> wrote:
>
> new data obtained by finding two books online:
Ok, this looks fine, and I'll happily update the source to use a quartic.
I did ask people about this on G+, and it turns out there's a few
people out there that seem to be knowledgeable. The suggestion was to
use a "virial series" which is really just he same thing as your
polynomial fit, but with the 0-order coefficient set to 1.0 exactly.
But the guy who knew the math says that our coefficients look good
too, and the fact that our 0-order coefficient is very close to 1.0
ends up being a verification that the least-squares fit did a good
job.
I'll update the numbers, re-run my comparison, and if the linear mix
now looks like it matches the air table better, I'll remove the
special case for air entirely.
Thanks,
Linus
_______________________________________________
subsurface mailing list
[email protected]
http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
