To get a random number

Hi,

   Am having a incremental column matrix like, e.g. for 1st iteration, it is 100*1(double), then for 2nd iteration iteration, it will be 100*2(double), etc.. Form this, for 1st iteration, i need to get a random no (single data only). from 1st column of matrix  and for 2nd iteration, i have to get a random no. from 2nd column of matrix only.

              Thank you in advance!!@!!

[Michael]

Maybe I misunderstand but to me this question appears contradictory. How is a number random if it depends on your column as an input?

Answer 1

[Wayne King]

You can use

x = randperm(100);
randomindex = x(1);

Then, use that index to choose an element from the appropriate column, say your data matrix is X

X(randomindex,1)  %choose from first column
X(randomindex,2)  %choose from 2nd column

[Muruganandham Subramanian]

Hi wayne,

Am having negative integers also.

[Wayne King]

How can you have negative indices in your vectors? that's not possible.

[Muruganandham Subramanian]

Hi wayne,

to get the vectors(i.e. 100*1,100*2), i used unifrnd command in the range b/w -0.8 to 0.2 for 100 populations. From this, i have to pick a data randomly.

[Walter Roberson]

The range of your data matrix X does not affect the use of Wayne's suggested code.

However, to clarify his code a little:

rchoices = randperm(size(X,2));

randfirst = X(rchoices(1),1);

rchoices = randperm(size(X,2));

randsecond = X(rchoices(1),2);

Repeating the randperm() is required if you want it to be possible for the same row to be chosen for both columns. If the same row must _not_ be chosen for the columns, then

rchoices = randperm(size(X,2));

randfirst = X(rchoices(1),1);

randsecond = X(rchoices(2),2);

[Muruganandham Subramanian]

Hi walter,

It's not working.

am having both negative and non-negative integers of 100*1, In this i have to pick single data randomly.

[Walter Roberson]

Sorry, the references to size(X,2) should have been size(X,1)

rchoices = randperm(size(X,1));

randchoice = X(rchoices(1),IterationNumber);

[Muruganandham Subramanian]

It's working...

Thanks walter