# What is a function that takes the diagonal of a matrix (N) and converts the diagonal to all zeros.

2 views (last 30 days)
Sean on 9 Apr 2013
Ex.
>> m=[1 2 3; 4 5 6;1 2 3]
m =
1 2 3
4 5 6
1 2 3
>> m=diagzero (m)
m =
0 2 3
4 0 6
1 2 0

Youssef Khmou on 9 Apr 2013
try :
function Y=diagzero(X)
N=size(X);
if N(1)~=N(2)
error(' Matrix not square');
end
Y=X;
for x=1:N(1)
Y(x,x)=0;
end

James Tursa on 10 Apr 2013
Edited: James Tursa on 10 Apr 2013
Another method if you know m is square:
r = size(m,1) + 1;
m(1:r:end) = 0;
If you don't know if m is square or not you could do this:
r = size(m,1);
n = min(numel(m),r*r);
m(1:r+1:n) = 0;

Andrei Bobrov on 10 Apr 2013
m(eye(size(m))>0) = 0;
James Tursa on 10 Apr 2013
Or to avoid testing all those 0's:
m(speye(size(m))>0) = 0;
or:
m(logical(speye(size(m)))) = 0;

Azzi Abdelmalek on 9 Apr 2013
Edited: Azzi Abdelmalek on 9 Apr 2013
ii=1:size(m,1);
m(sub2ind(size(m),ii,ii))=0

Jonathan Epperl on 10 Apr 2013
Edited: Jonathan Epperl on 10 Apr 2013
Assuming m is square (because else you'd have to define the diagonal for me):
diagzero = @(m) m - diag(diag(m));
James Tursa on 10 Apr 2013
Doesn't work on any diagonal elements that are inf or nan.