Jos

It is nice to see some improvments Franciso! A couple of additional comments: if your sor worried about speed, I suggest you take a look at logical indexing, and replace all the zeros from the struct command by one call to zeros.

Another point: the behavior for 2D matrices should be explained, especially when the groups are specified. I think you did not intend that behavior. For instance, take a look if the following produces what you want them to produce:

A = rand(3,4) ; allstats(A), allstats(A,1:4), allstats(A,1:3)

With respect to the help, you can add examples of the expected output, besides giving only the calling syntax.

Below is the outline of a more optimized code:
function R = allstats(DATA, GROUP)
if isempty(ver('stats'))
    error('Function requires STATS toolbox.') ;
end
if nargin==1,
    GROUP = ones(numel(DATA),1) ;
    Ngroups = 1 ;
    UniqueGroups = 1 ;
else
    if numel(DATA) ~= numel(GROUP)
        error('Number of elements should match') ;
    end
    UniqueGroups = unique(GROUP) ;
    Ngroups = numel(UniqueGroups) ;
end
Z = nan(Ngroups,1) ;
R = struct('min',Z,'max',Z,'mean',Z) ; % etce
for k=1:Ngroups,
    Q = GROUP == UniqueGroups(k) ; % logical indexing
    R(k).min = nanmin(DATA(Q)) ;
    R(k).max = nanmax(DATA(Q)) ;
    R(k).mean = nanmean(DATA(Q)) ;
end

Here are the first few lines of this script:
%Cronometro.
clc
clear
resp='s';
while (resp=='s')

clc
clear
a=input('Ingrese horas\n\n');
disp(' ')
b=input('Ingrese minutos\n\n');
disp(' ')
c=input('Ingrese segundos\n\n');
disp(' ')

Potential users should be aware that running this script does clear the entire workspace without notice (and it do so twice!), so all your work will be lost.
This script serves no purpose other than being a programming exercise for the author, so why share it here?

Just what I was looking for

I've tried this small experiment:

N = 100000; % sample size
w = 1:100; % quite steep distribution
w = w / sum(w); % distribution (pdf)
Y_std = sqrt(w .* (1 - w)); % Standard deviations associated to each probability

%%%%%%%%%%%%%%%%%

Y1 = randp(w, N, 1); % Try Jos function
Y1_w = histc(Y1, [1:100]); % absolute frequencies table
Y1_w = Y1_w / N; % transform into relative frequencies
Y1_w_std = sqrt(Y1_w .* (1 - Y1_w)); % Empirical standard deviations associated to each probability

%%%%%%%%%%%%%%%%%

Y2 = gDiscrPdfRnd(w, N, 1); % Try Gianluca Dorini's function
Y2_w = histc(Y2, [1:100]); % absolute frequencies table
Y2_w = Y2_w / N; % transform into relative frequencies
Y2_w_std = sqrt(Y2_w .* (1 - Y2_w)); % Empirical standard deviations associated to each probability

%%%%%%%%%%%%%%%%%

% Check if empirical standard deviations are coherent with theretical ones
plot(Y_std - Y1_w_std', '.b') % plot empirical differences for Jos function
hold on
plot(Y_std - Y2_w_std', '.r') % plot empirical differences for Gianluca Dorini's function

%%%%%%%%%%%%%%%%%

It looks as though both functions statistically behave very well. They are both excellent functions. Thank you.

Wow, it really is a nice function, very fast and efficient. It is exactly what I was looking for. Thank you.