what will be the solution for this error

function idx=nomc(meas) 
W=SimilarMatrix(meas,1); 
D=diag(sum(W)); 
L=D-W; 
Lsym=(D^(-1/2))*L*(D^(-1/2)); 
[ev ed]=eig(Lsym); 
d=diag(ed);  %#ok<NASGU>
V=ev(:,1:3); 
for i=1:250 
    for j=1:7 
        U(i,j)=V(i,j)/sqrt(sum(V(i,:).^2));  %#ok<AGROW>
    end 
end 
idx=kmeans(U,3); 
figure;plot(idx); 
result=[length(find(idx==1)),length(find(idx==2)),length(find(idx==3))];

getting error

??? Attempted to access V(1,4); index out of bounds because

size(V)=[143,3].

Error in ==> nomc at 11

        U(i,j)=V(i,j)/sqrt(sum(V(i,:).^2));  %#ok<AGROW>

Error in ==> Normalizemeanseg at 120

f=nomc(y);

[Rushikesh Tade]

can you provide "meas" variable in mat format or provide dimensions?

[Poonam]

 meas is <143x256x3 double>

Answer 1

[Titus Edelhofer]

Hi,

you write

V = ev(:, 1:3);

so V has three columns. But your loop on j is

for j=1:7

That doesn't match ...

Titus

[Poonam]

if i write

for j=1:3

still giving same error

[Pierre Benoit]

Well you should check the value of i then. You should use the sizes of V for your loop boundaries rather than some magic numbers.

Type

dbstop if error

in your command window before running your code, it will be easier to understand what went wrong.

[Image Analyst]

Poonam, I think you're mistaken. If the error says

Attempted to access V(1,4);

and your code is

for j=1:3

Then there is no way the second index of V could be 4 in this line

U(i,j)=V(i,j).....

[Pierre Benoit]

Well, my guess was that this is no longer j that was causing a problem but i since the boundary of this loop is also quite arbitrary. And, maybe the author forgot to tell us or didn't see that the error changed a little.

But at this point, we can only guess given how little information we have.