could anyone help me to solve the issue

I am having a matrix size(4,8) which is having a combination of zero and non zero values.

could any one help me how to display the matrix in such a way that it should atleast contain one nonzero value with respect to its every row and every column

[Rik]

Are there any further restrictions? What values should be put in?

[Walter Roberson]

This looks to me as if it is a duplicate of https://www.mathworks.com/matlabcentral/answers/465579-could-anyone-help-me-to-solve-the-issue-in-the-following-code

[jaah navi]

I am having the matrix

a=[ 0 0 0 0 5.8209 0 0 0;

0 0 6.0208 0 0 0 0 0;

6.3376 5.3833 0 6.4611 6.0262 6.2584 5.6534 6.4254;

5.4371 5.2587 6.2028 5.9251 0 5.9764 6.0064 5.9132]

I need to modify the above matrix in such a way that every row and column should contain atleast one non zero value.

[jaah navi]

I need to get the result in the following manner

b=[ 0 0 0 0 5.8209 0 0 0;

0 0 6.0208 0 0 0 0 0;

6.3376 0 0 6.4611 0 0 5.6534 0;

0 5.2587 0 0 0 5.9764 0 5.9132]

in this result i can get atleat one non zero value with respect to every column and every row.

[Rik]

Does it matter which values are removed? What motivates the choice? Should there be as little non-zero values left as possible?

[Matt J]

I need to modify the above matrix in such a way that every row and column should contain atleast one non zero value.

The original matrix you showed already did satisfy this requirement.

[jaah navi]

the values can be removed under the condition that every row and column should contain atleast one non zero values.

Answer 1

[Rik]

This removes elements until your matrix is no longer valid according to your rule. It is very slow for bigger matrices, but it is guaranteed to work. Look up "bubble sort" if you want to understand how it works.

A=rand(4,8);

isValidMatrix=@(x) all(any(x,1)) && all(any(x,2));

B=A;
%use a structure similar to the 'bubble sort' algorithm
for idx1=1:numel(B)
    for idx2=idx1:numel(B)
        if B(idx2)==0,continue,end
        B_temp=B;
        B_temp(idx2)=0;
        if isValidMatrix(B_temp)
            B=B_temp;
        end
    end
end
disp(B)