Thread Subject: combinations of N matrixes taken K at a time

Hey I’m a new user.
I want to perform this:

I have 6 square matrixes A B C D E F
I want to find the possible combinations of these matrixes taken 3 at a time and calculate the product of the determiner of each combination
In order to find all possible combinations I can do this:
MAT={‘A’,’B’,’C’,’D’,’E’,’F’}
comb=nchoosek(MAT,3)
The results is:

comb =

    'A' 'B' 'C'
    'A' 'B' 'D'
    'A' 'B' 'E'
    'A' 'B' 'F'
    'A' 'C' 'D'
    'A' 'C' 'E'
    'A' 'C' 'F'
    'A' 'D' 'E'
    'A' 'D' 'F'
    'A' 'E' 'F'
    'B' 'C' 'D'
    'B' 'C' 'E'
    'B' 'C' 'F'
    'B' 'D' 'E'
    'B' 'D' 'F'
    'B' 'E' 'F'
    'C' 'D' 'E'
    'C' 'D' 'F'
    'C' 'E' 'F'
    'D' 'E' 'F'
Then I know all combinations

Now I want to create with a for cycle a matrix with like this:

Det=[|A|*|B|*|C| , | A|*|B|*D|,………]
How can I perform this multiplication?
The problem is that in the matrix comb, A is a text.
Please give me some advice
Thanks

Store your matrices in a cell array, then use nchoosek with numeric arguments to index into the cell.

A = {rand(2) rand(2) rand(2)}
t = nchoosek(1:3,2)

G = cell(size(t,1),1);
for ii = 1:size(t,1)
    G = A(t,ii,1) * A(t,ii,2); % Or whatever here.
end

Sorry, that should read:


A = {rand(2) rand(2) rand(2)}
t = nchoosek(1:3,2)

G = cell(size(t,1),1);
for ii = 1:size(t,1)
    G{ii} = A{t(ii,1)} * A{t(ii,2)}; % Or whatever here.
end

"tirinacria CICCIOCICCIO" <d.accorso@tin.it> wrote in message <hat0le$jeq$1@fred.mathworks.com>...

> The problem is that in the matrix comb, A is a text.

When you call nchoosek, you can supply any kind of arrays or cell arrays. Example:

A = rand(5);
B = rand(5);
C = rand(5);
D = rand(5);
E = rand(5);
F = rand(5);

allmystuff = {A B C D E F};

threestuff = nchoosek(allmystuff, 3);

% Each row of the result THREESTUFF is one combination

If you want to compute the product of determinants, then feed NCHOOSEK with the array of determinants of A, B, ..., F, then compute PROD of the result along the second dimension.

Bruno

thank you very much!