A Random Question

[Joe Horn]

(06-09-2017 03:15 PM)DavidM Wrote:  ... I'm wondering if the second approach reduces the quality of the randomization in any way. My inclination is to believe that the second method should be at least as good as (and possibly better than) the first, but I know better than to rely on a gut feel for this sort of thing. Are there problems with the second approach that would weaken the randomization when compared to the first?

My strong gut feeling is that using MANT on random X's is a Bad Thing, because it SEEMS like it simply multiplies X by 10, but that's only if X>=0.1, which it isn't 10% of the time. And 1% of the time X<0.01, which is even worse. And 0.1% of the time X<0.001, et cetera. In effect, MANT(X) where X=RAND returns X*10^(-XPON(X)). This will not return uniformly random numbers at all.

Quote:How could I test this?

I'd suggest writing a loop on the HP 50g which turns on random pixels on the screen until you abort it by pressing a key. Write one version of the program using (RAND,RAND) scaled to the screen, to see what it should look like. Then write another version using (MANT,MANT) scaled to the screen. I'll bet a wooden nickel that you'll see a lack of uniformity quite clearly. I haven't tried it, so I might be wrong, in which case I'll have to go find a wooden nickel.