On 11/18/2014 08:40 AM, andy pugh wrote: > On 18 November 2014 15:32, dave <[email protected]> wrote: >> On another tangent I wonder about using a polynomial (curve fit) but >> with the number of inflections points you have it may be pretty high >> order. Linear interp instead of a continuous function may still be best. >> This is the kind of thing that starts email wars. ;-) > It is a constant source of friction between US and UK teams in my job. > The US like models and polynomials, we like lookup tables as they are > faster, can have discontinuities, and behave a lot better > out-of-range. > Andy, Can't argue with your statements. Polynomials are unbeatable where the function and the real world match well. Using polynomials out of range is not only foolish but also can be dangerous. The ideal here would be to fix the surface but that may not be or even possible unless one just throws money at it. If you don't need a continuous function and lookup will do the job it is probably better.
------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
