Thread Subject:

array with random not repeated numbers

hi,

how can I generate an array with random numbers (given a min and max) ?
Important: they have to be not repeated.

thanks

On Nov 10, 5:04 pm, "Patrick Diviacco" <patrick.divia...@gmail.com>
wrote:
> hi,
>
> how can I generate an array with random numbers (given a min and max) ?
> Important: they have to be not repeated.
>
> thanks

-----------------------------------------------------------
Did you see the rand() function?
So how about something like this
randomArray = minValue + (maxValue-MinValue)*rand(rows, columns);

On Nov 10, 3:37 pm, ImageAnalyst <imageanal...@mailinator.com> wrote:
> On Nov 10, 5:04 pm, "Patrick Diviacco" <patrick.divia...@gmail.com>
> wrote:
>
> > hi,
>
> > how can I generate an array with random numbers (given a min and max) ?
> > Important: they have to be not repeated.
>
> > thanks
>
> -----------------------------------------------------------
> Did you see the rand() function?
> So how about something like this
> randomArray = minValue + (maxValue-MinValue)*rand(rows, columns);

Does that ensure that there won't be any repeated values though?

ImageAnalyst <imageanalyst@mailinator.com> wrote in message <45a8dd1f-8806-4903-ac8f-6a25382539f0@j19g2000vbi.googlegroups.com>...
> On Nov 10, 5:04?pm, "Patrick Diviacco" <patrick.divia...@gmail.com>
> wrote:
> > hi,
> >
> > how can I generate an array with random numbers (given a min and max) ?
> > Important: they have to be not repeated.
> >
> > thanks
>
> -----------------------------------------------------------
> Did you see the rand() function?
> So how about something like this
> randomArray = minValue + (maxValue-MinValue)*rand(rows, columns);


Is this code avoiding duplicates ?

Patrick Diviacco wrote:
> ImageAnalyst <imageanalyst@mailinator.com> wrote in message <45a8dd1f-8806-4903-ac8f-6a25382539f0@j19g2000vbi.googlegroups.com>...
>> On Nov 10, 5:04?pm, "Patrick Diviacco" <patrick.divia...@gmail.com>
>> wrote:
>>> hi,
>>>
>>> how can I generate an array with random numbers (given a min and max) ?
>>> Important: they have to be not repeated.
>>>
>>> thanks
>> -----------------------------------------------------------
>> Did you see the rand() function?
>> So how about something like this
>> randomArray = minValue + (maxValue-MinValue)*rand(rows, columns);
>
>
> Is this code avoiding duplicates ?

No, but odds are pretty good unless the number you're asking for is very
high or the the range very constricted.

You can, of course, run a rejection/regeneration phase after the initial
generation.

--

"Patrick Diviacco" wrote in message <hdco0i$1kk$1@fred.mathworks.com>...
> hi,
>
> how can I generate an array with random numbers (given a min and max) ?
> Important: they have to be not repeated.
>
> thanks

Do you mean integer numbers
for example if the number is from min=1 to max= 9

we need to generate array of length n=5
min=1
max=9
A=min:max
x=zeros(1,n)
for k=1:n
     x(k) = randsrc(1,1,A)
     a= find(A==x(k))
     A(a)=[];
end

"Namo " <namo1983@hotmail.com> wrote in message <isg2tg$fvr$1@newscl01ah.mathworks.com>...
> "Patrick Diviacco" wrote in message <hdco0i$1kk$1@fred.mathworks.com>...
> > hi,
> >
> > how can I generate an array with random numbers (given a min and max) ?
> > Important: they have to be not repeated.
> >
> > thanks
>
> Do you mean integer numbers
> for example if the number is from min=1 to max= 9
>
> we need to generate array of length n=5
> min=1
> max=9
> A=min:max
> x=zeros(1,n)
> for k=1:n
> x(k) = randsrc(1,1,A)
> a= find(A==x(k))
> A(a)=[];
> end
- - - - - - - - - -
  For n random non-repeating integers in [min:max] just use randperm:

 p = randperm(max-min+1);
 p = p(1:n)+min-1;

Roger Stafford

"Patrick Diviacco" wrote in message <hdco0i$1kk$1@fred.mathworks.com>...
> hi,
>
> how can I generate an array with random numbers (given a min and max) ?
> Important: they have to be not repeated.
>
> thanks


See this thread:

http://www.mathworks.com/matlabcentral/newsreader/view_thread/304272#824982

Derek O'Connor

"Derek O'Connor" wrote in message <isjc8p$j7j$1@newscl01ah.mathworks.com>...
> See this thread:
>
> http://www.mathworks.com/matlabcentral/newsreader/view_thread/304272#824982
>
> Derek O'Connor
- - - - - - - - - - -
  Thanks for reminding us of that thread, Derek. I had completely forgotten about that interesting discussion.

  This is an alternate method for choosing n non-repeating random integers out of [min:max] that avoids 'randperm' and its sorting:

 p = min:max;
 m = length(p);
 r = zeros(1,n);
 for k = m:-1:m-n+1
  q = ceil(k*rand);
  r((m+1)-k) = p(q); % <-- n random integers go to 'r'
  p(q) = p(k);
 end

  This method might be preferable in cases where max-min is so large as to require excessive time be spent sorting in 'randperm'.

Roger Stafford

"Roger Stafford" wrote in message <isju4k$6ok$1@newscl01ah.mathworks.com>...
> "Derek O'Connor" wrote in message <isjc8p$j7j$1@newscl01ah.mathworks.com>...
> > See this thread:
> >
> > http://www.mathworks.com/matlabcentral/newsreader/view_thread/304272#824982
> >
> > Derek O'Connor
> - - - - - - - - - - -
> Thanks for reminding us of that thread, Derek. I had completely forgotten about that interesting discussion.
>
> This is an alternate method for choosing n non-repeating random integers out of [min:max] that avoids 'randperm' and its sorting:
>
> p = min:max;
> m = length(p);
> r = zeros(1,n);
> for k = m:-1:m-n+1
> q = ceil(k*rand);
> r((m+1)-k) = p(q); % <-- n random integers go to 'r'
> p(q) = p(k);
> end
>
> This method might be preferable in cases where max-min is so large as to require excessive time be spent sorting in 'randperm'.
>
> Roger Stafford

------------------------------------

Roger,

I think your method is a modification of the Fisher-Yates shuffle
algorithm as implemented by Durstenfeld, CACM 1964. It is nearly
always preferable to Matlab's randperm which is O(n log n) because
it must sort. If speed is critical then Jan Simon's mexed Shuffle could
used:

http://www.mathworks.com/matlabcentral/fileexchange/27076-shuffle

If the sample size n is very much less the range N = max-min+1,
then the following function is fast and uses only O(n) memory.

%--------------------------------------------------------------
function p = RandSampNR(min,max,n);
% -------------------------------------------------------------
% Generate a random sample of size n from the range of N integers
% min:max, without replacement, using a rejection loop.
% Time Complexity is at least O(n^2) and is suitable for small
% n << N = max-min+1 only. Note that this is a Las Vegas algorithm
% whose running time is random, because of the rejection while-loop.
% Space Complexity is O(n).
%
% Derek O'Connor, 12 Mar 2011. derekroconnor@eircom.net
%
% USE: p = RandSampNR(-2^29,2^29,10^3)
% -------------------------------------------------------------
p = zeros(1,n);
N = max-min+1;
p(1) = min + floor(rand*N);
for k = 2:n
      r = min + floor(rand*N);
      while member(r,p,k-1)
           r = min + floor(rand*N);
      end
      p(k) = r;
end

function answer = member(e,v,k)
answer = false;
for i = 1:k
    if v(i) == e
        answer = true;
        return
    end
end

%---------------------------------------------------------------

Example:

>> tic;r = RandSampNR(-2^29,2^29,10^3);toc
Elapsed time is 0.012196 seconds.
Memory: about 8 Kbytes

>> tic;r = RandSampNRStafford2(-2^29,2^29,10^3);toc
Elapsed time is 10.120162 seconds.
Memory: about 8 Gbytes

Dell Precision 690, Intel 2xQuad-Core E5345 @ 2.33GHz 16GB RAM,
Windows7 64-bit Professional. MATLAB Version 7.6.0.324 (R2008a)


Derek O'Connor