Re: Testable Predictions by HdB
*** T. Winter wrote:
You misunderstand the word stepsize. Stepsize is the distance between
successive samples. However, your article mentions nothing about all
that information (at least 25 points, etc.).
That's true. The article is incomplete. But I can't help it right now.
> > Decreasing h beyond that figure indeed had bad effects on the precision.
> > Just as I "predicted".
>
> Worse. It's even possible to have an "underflow" with these Gaussians.
But originally you denied that you could make the stepsize too small.
Not if you keep it sensible. I defined kind of a "standard error" by
exp(-(2.pi)^2/2) for that purpose, giving 25 samples and a step size
which is not too small. Geez, is that so difficult ? But I guess it's
compiler-and-so-on dependent. Tried exp(-1.E+10000) and finally did
succeed in Halting the program. (That's _much_ later than I expected)
Oh, well, I currently do not have access to a Pascal compiler.
I currently do not have access to a C compiler. But that's just a minor
hurdle. I can easily translate a page or two in my favorite speech.
Han de Bruijn
.
