Re: Geometrically distributed random numbers on Rabbit 2000.

On Wed, 08 Jun 2005 09:30:11 GMT, Jonathan Kirwan
<jkirwan@xxxxxxxxxxxxxx> wrote:

>On Wed, 08 Jun 2005 18:08:22 +1000, The Real Andy
><will_get_back_to_you_on_This> wrote:
>
>>So me thinks more, i wonder if its possible to somehow converge the
>>numbers to speed things up.
>
>Knuth's algorithm is:
>
> INT( LN( uniform ) / LN( 1 - p ) )
>
>where 'uniform' is a uniform deviate and p is the probability. Knuth
>mentions this on page 131 in Seminumerical Algorthms. But you already
>know that, I assume, since you mentioned it.
>
>Are you only looking for an algorithm that will iterate geometrically?
>If so, Knuth discusses, in exercise 6 on page 133, such an algorithm.
>
>Other than that, I don't have a clue about what you want.
>
>Jon

I have not got that book. I must go buy it.

The algorithm you mention above I had just located about half an hour
ago on the internet. I am currently running tests of it (not on the
rabbit however) to see if it suits my needs

The knuth algorith I have mentioned is for a uniformly distributed
RNG. It was already approved by the government regulator for use in
gaming applications, hence the reason for using it.

Many thanks.

.

Relevant Pages

  • Re: Geometrically distributed random numbers on Rabbit 2000.
    ... If you test the algorithm that Westhues and I mention, ... You imagine a circle centered inside a square and sized ... get two random uniform deviates U and V, ...
    (sci.electronics.design)
  • Re: -- random points on the unit n-sphere
    ... I came up with one algorithm on my own, and I implemented it in Maple. ... Surprising fact that this distribution is in fact uniform follows from ...
    (sci.math)
  • Re: -- random points on the unit n-sphere
    ... I came up with one algorithm on my own, and I implemented it in Maple. ... Surprising fact that this distribution is in fact uniform follows from ... Here's my Mathematica implementation ...
    (sci.math)
  • Re: Generation of range permutations?
    ... > where uniform() can be implemented using Benjamin Goldberg's ... it's not "my" algorithm; ... I recently discovered that the divide and conquer shuffle algorithm ... which I had thought was my own invention was previously discovered on ...
    (sci.crypt)
  • Re: Geometrically distributed random numbers on Rabbit 2000.
    ... i wonder if its possible to somehow converge the ... where 'uniform' is a uniform deviate and p is the probability. ... Are you only looking for an algorithm that will iterate geometrically? ... If so, Knuth discusses, in exercise 6 on page 133, such an algorithm. ...
    (sci.electronics.design)