why the error is coming?

2 views (last 30 days)
VIJENDRA
VIJENDRA on 24 Jun 2014
Trying to execute this code but it shows error "Subscript indices must either be real positive integers or logicals."
[M,N]=size(x); y=zeros(1,N); for l=1:N sum1=sum(x(:,l).*log(x(:,l))); sum1=-1*sum1; y(1,l)=sum1; end

Accepted Answer

David Sanchez
David Sanchez on 24 Jun 2014
who is x?
That message is the result of trying to access an non-existent element of the array, such as x(0).
Make sure x has the right size.
By the way, your for-loop code
x= rand(3,4); % sample data
[M,N]=size(x);
y=zeros(1,N);
for l=1:N
sum1=sum(x(:,l).*log(x(:,l)));
sum1=-1*sum1;
y(1,l)=sum1;
end
Is the same that this, in a single line:
for l=1:N
y(l) = -( sum(x(:,l).*log(x(:,l))) );
end
  1 Comment
VIJENDRA
VIJENDRA on 24 Jun 2014
x is a matrix and still showing this same error.

Sign in to comment.

More Answers (1)

Roger Stafford
Roger Stafford on 24 Jun 2014
Edited: Roger Stafford on 24 Jun 2014
It is likely that somewhere in your system you have defined a variable named either 'sum', 'log', 'size', or 'zeros'. If so, this would be in conflict with one of matlab's functions with the same name and this would give rise to the error message you are receiving.
You can also write it as:
y = sum(-x.*log(x),1);
  1 Comment
Alfonso Nieto-Castanon
Alfonso Nieto-Castanon on 24 Jun 2014
my money is on sum being a variable, given the choice of variable names used in the example...
clear sum;
should fix this issue...

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!