Delete and merge rows and columns based on values of other matrix under looop

i have a matrix of A and B1 B2 B3....Bn condition is to check B1 if zero then eliminate the rows and columns in matrix A store as A1 then it should go to B2 check and store in A2 till An.

A is nxn matrix

B1 B2.....Bn is a 1xn

Example:

if any of the element in B1 matrix is zero suppose element 3 it should eliminate 3rd row and 3rd column of Matrix A and store as A1

and it has to loop till Bn and store An matrices.

[Turlough Hughes]

Very doable, but I would first suggest grouping B1 to Bn into an n by n matrix and grouping your results A1 to An into probably a cell array, if possible. At the very least you could then write a for loop to iterate the answer rather than writting out the same expressions n times.

[vimal r]

share the code for Bnxn and Anxn .

[Turlough Hughes]

You could do something like this...

% Generated a random array A and array B where approx 20% of the values in B are 0
A = rand(100);
Bn = rand(100)>0.2;

Similary to your description, column 1 of B corresponds to B1, column 2 corresponds to B2, etc. To remove the rows and columns corresponding to 0's in each column you could then do the following:

An = cell(1,size(A,1)); % Preallocate An
for c = 1:size(A,1)
    temp = A;
    temp(~Bn(:,c),:) = [];
    temp(:,~Bn(:,c)) = [];
    An{1,c} = temp;
end

[vimal r]

i have to store An matrix loop should end n variable output of A1 ....An based on the condition of Bnxn each row as one senario...

my input is Anxn and Bnxn

take first row of B check zeros and based on that it should eliminate rows and columns in matrix A and store as A1

then go to second row of B check and based on zero eliminate rows and columns in matrix A and store as A2

so on.....

[Turlough Hughes]

The code I provided above does what you ask with the exception of working through columns of B as opposed to rows of B because I did not have enough info at the time... that's is a simple fix. Having variables numbered A1 to An is a sign that you are doing something wrong because you either have to copy and paste code or use the eval function to generate these variables.

https://uk.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

The results in the example I provided above are stored in a cell array which is in my opinion the easiest alternative to generating n variables. The results are then equivalently stored as

An{1,1}, An{1,2} all the way up to An{1,n}.