random-linear: Linearly distributed in [0,1) with a mean value of
0.2929.  The density function is given by 'f(x) = 2 * (1 - x)'.

random-inverse-linear: Linearly distributed in [0,1) with a mean
value of 0.6969.  The density function is given by 'f(x) = 2 * (x -
1)'.

random-triangular: Triangularly distributed in [0,1) with a mean value of 0.5.

Exponentialy distributed with a mean value of '0.69315 / l'.  There
is no upper limit on the value however there is only a one in
one-thousand chance of generating a number greater than '6.9078 /
l'.  The density function is given by: 'f(x) = l ^ (-l * x)'.

Bilinear exponentialy distributed with a mean value of 0 and where
half of the results lie between '+- 1 / l'.  The density function
is given by: 'f(x) = 0.5 * l * (e ^ (-l * |x|))'.

Guassian distributed with a mean value of `mu' and where 68.26% of
values will occur within the interval +-`sigma' and 99.75% within
the interval +-(3 * `sigma').  The density function is given by:
'f(x) = (1 / SQRT (2 * pi * sigma)) EXP - ((x - u) ^ 2) / (s *
(sigma ^ 2))'.

Cauchy-distributed with a mean value of 0 where half of the results
lie in the interval +-alpha, and 99.9% fall within +-318.3alpha.
The density function is given by: 'f(x) = alpha / (pi * (alpha ^ 2
+ x ^ 2))'.

Beta-distributed in [0,1).

Random permutation.  For a critique of this method see
http://okmij.org/ftp/Haskell/perfect-shuffle.txt