Re: generating unique random numbers

a kusa schrieb am 22.02.2010 um 11:14:11 (-0600):

> What I did was to call math:random() twice and adding the results each
> time around.

I don't think addition significantly lowers the likelihood of getting
two equal numbers. Imagine rolling dice, there is a chance of 1:6 of
getting two equal numbers. Now take two dice, and try to match the sum
of the first round in the second round: the probability will depend on
the result of the first round and will be bell-curved, meaning higher
probability for values in the middle and a value-space from 2 to 12.

Now imagine multiplying the results: The value-space becomes largesr,
sparser, and hence less prone to duplicates.

Do the Perl:

perl -lwe 'for $x(1..6) {for $y(1..6){print$x+$y}}' | sort -n | uniq -c 
      1 2
      2 3
      3 4
      4 5
      5 6
      6 7
      5 8
      4 9
      3 10
      2 11
      1 12

perl -lwe 'for $x(1..6) {for $y(1..6){print$x*$y}}' | sort -n | uniq -c 
      1 1
      2 2
      2 3
      3 4
      2 5
      4 6
      2 8
      1 9
      2 10
      4 12
      2 15
      1 16
      2 18
      2 20
      2 24
      1 25
      2 30
      1 36

That's not a proof, of course, it's just the helpless expression of what
I think is the case here. More knowledgeable mathematical assessments
appreciated!

> And I used distinct-values() everytime.

How do you use it?

-- 
Michael Ludwig