Jos (10584)

The help is certainly an improvement! You could add an H1 line to assist users of LOOKFOR.

However, floating point issues are not dealt with properly : rand_gen(x,[.1 .2 .3 .3 .1],N) returns an unexpected error .

And, as Michael Chen already pointed out, there are many alternatives on the FEX, including my own:
http://www.mathworks.com/matlabcentral/fileexchange/8891

How is this different from the more simple approach:

idx = ceil(numel(alpha)*rand(n,w)) ;
words = alpha(idx)?

Note that the variable "word" is growing inside the lloop, making this algorithm potentially very slow.

A final and positive comment: the help is quite adequate with a description of the in- and outputs, the algorithm and an example. A "See Also" line, referencing RAND and RANDSAMPLE would improve it even further.

Nicely coded, with copious internal comments! As a suggestion for improvement, you could mention in the help that the coefficient can range from 0 (no dispersion, complete equality) to 1 (total inequality), as you do on the fex page. By the way, ginicoeff(ScalarValue) is now 0, but I think it should be 1, as it is total inequality since one gets all, but it is also equal ... I got confused.
And why is the correction not the default, so it truly ranges from 0 to 1?

I should also add that this function doesn't wait for input. So if no keys are pressed when it is called, it is bypassed, which is great for use in a while loop.

@ J B. Please see my comment of 10 dec 2009:

h = errorbar(x,y,ex) ;
set(h,'linewidth',2)