how to create 200 random matrices using loop..

Hi..I want to create separate 200 random matrices of size 32*32 , and each to be multiplied with my input matrix..whether any looping is possible here ..please help ..

thank in advance

Answer 1

[Image Analyst]

Try this:

tic;
inputMatrix = rand(32, 32); % Whatever you want...
for k = 1 : 200
  thisRandomMatrix = rand(32,32); % Create new random matrix for this iteration.
  storedMatrixes{k} = inputMatrix .* thisRandomMatrix ; % Multiply them
end
toc;
msgbox('Done with loop');

[Abhishek sadasivan]

I would like to get separate 32 output matrices after multiplication so that I can take the sum of all elements in each matrix (I am trying to project my input data into random matrices ).Each sum will be one measurement ..

[Image Analyst]

They are separate. Each one is in its own cell. To sum up all the elements in a particular matrix you can extract it then sum it:

thisMatrix = storedMatrixes{10}; % Get matrix #10.
theSum = sum(thisMatrix(:));  % Sum of all elements of matrix #10.

Read the FAQ on cell arrays: http://matlab.wikia.com/wiki/FAQ#What_is_a_cell_array.3F

You could also store them as planes/slices in a 3D matrix if you preallocate a 3D matrix for them.

Answer 2

[Matz Johansson Bergström]

I have not tried it but something like this:

n=200;
tmp = rand(32,32,n);

and then to multiply each matrix

for i=1:n
in = in*tmp(:,:,i)
end

[Abhishek sadasivan]

thanks ..but I would like to get separate 32 output matrices after multiplication so that I can take the sum of all elements in each matrix (I am trying to project my input data into random matrices ).Each sum will be one measurement ..

[Matz Johansson Bergström]

I would argue that this is possible and very simple using a 3D matrix. The data is stored in another way, that's all and very efficient, because Matlab is made for matrices. For small matrices this does not really matter, but for larger ones it might make a difference.

My suggestion stacks the matrices into a 3d matrix and you can simply pick out the matrices like you would any other data structure.

Also, to sum all the matrices, you just have to write

sums = squeeze(sum(sum(tmp)))

This gives you the sum of all the elements in each matrix as a vector. Sums(i) contains the sum of matrix i.