🚀 go-pugleaf

In article <C1FpMw.4Ip@ccu.umanitoba.ca> sbloch@silver.cs.umanitoba.ca (Stephen Bloch) writes:
>sumner@math.scarolina.edu (David Sumner) writes:
>>> In an earlier post I wrote that an easy way to generate a
>>> random permutation of the integers 1-n was:
>>> >      ++++++++++++++
>>> >  Initialize  A[n] as the array [1, 2, 3, 4,..., n]
>>> >        for i=1 to n
>>> >        z = random(n)
>>> >        t = a[i]
>>> >        a[i] = a[z]
>>> >        a[z] = t
>>> >        next i
>>> >    ++++++++++++++
>>> There is a typo in this. The line 'z = random(n)' should
>>> be 'z = random(i)'
>
>and "Thomas W. Strong, Jr." <strong+@CMU.EDU> replies:
>>I have to agree with the algorithm as originally presented.  What it
>>effectively does is it takes each item and puts it in a random position.
>><stuff deleted>

>No, the original algorithm is incorrect (for the simple combinatorial
>reason several people have posted).  

This is counterintuitive to a degree that it reminds me of the famous
"Monte Hall" flap of a couple years back.  There must be some way to
make money from this... ;^)

Joe N.