Create a diagonal matrix with a for loop from a vector

I want to use the ML vector to create a diagonal matrix of only the values of the ML vector on the diagonal and make a new ML matrix with zeros everywhere else and the values of the ML vector along the diagonal of the new ML matrix. Essentially I am trying to write the code for diag(ML).

ML = rand([5 1])

for j = 1:length(ML)

for i = 1:j

if i < j

ML(i,j) == 0

elseif i > j

ML(i,j) == 0

else

ML(i,j) = ML(j)

end

end

end

Answer 1

[Dyuman Joshi]

(If you are required to use a loop)

It is better to use another variable to get an output.

ML = rand([5 1])

ML = 5×1

0.7725 0.4016 0.8767 0.9691 0.1817

Pre-allocate output matrix according to the size.

%Using max() in case ML is a row vector
Mout = zeros(max(size(ML)))

Mout = 5×5

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

%You can also obtain the result without double for loop
for j = 1:numel(ML)
    %Directly define the values
    Mout(j,j)=ML(j);
end
Mout

Mout = 5×5

0.7725 0 0 0 0 0 0.4016 0 0 0 0 0 0.8767 0 0 0 0 0 0.9691 0 0 0 0 0 0.1817

Answer 2

[Adam Danz]

Here are three way to define a diagonal in a square matrix of zeros that do not require a loop.

eye()

ML = rand([5 1])

ML = 5×1

0.2650 0.0656 0.8188 0.8268 0.6532

Mout = eye(numel(ML)).*ML

Mout = 5×5

0.2650 0 0 0 0 0 0.0656 0 0 0 0 0 0.8188 0 0 0 0 0 0.8268 0 0 0 0 0 0.6532

Indexing with sub2ind

n = numel(ML); 
Mout = zeros(n); 
ind = sub2ind([n,n],1:n,1:n);
Mout(ind) = ML

Mout = 5×5

0.2650 0 0 0 0 0 0.0656 0 0 0 0 0 0.8188 0 0 0 0 0 0.8268 0 0 0 0 0 0.6532

Indexing without sub2ind

This shortcut works because we're working with a square matrix.

n = numel(ML); 
Mout = zeros(n); 
Mout(1:n+1:end) = ML

Mout = 5×5

0.2650 0 0 0 0 0 0.0656 0 0 0 0 0 0.8188 0 0 0 0 0 0.8268 0 0 0 0 0 0.6532

[Dyuman Joshi]

I wonder if this is also the implementation in diag(input) or diag(input,0)