1. Re: exp function
Jeff Zeitlin writes:
>>> Yes, I know I can approximate it with 2.718281828459045, but...
> Well, yes, I had gotten that far, and I suspect that my
> approximation will be more than accurate enough for ordinary use
> - the point I was indirectly making was that exp() is usually a
> built-in, and for all I know, it may _compute_ exp(1) rather than
> having it hardcoded in there as a constant.
Your approximation of e is 16 digits long.
That's roughly the limit of accuracy of the
floating-point hardware on Intel machines anyway,
so a computed value of e could not be much, if any,
more accurate.
Regards,
Rob Craig
Rapid Deployment Software
