Question about imresize?

Hi, I have this question:

I have a matrix whose pixel value represent a gaussian beam. For my experimental purpose I want to make it smaller in a matrix of equal dimensions and I have decided to do this:

 Q=padarray(A, [1000*t 1000*t], 0);
 SS=imresize(Q,[1000 1000]);

 

I encrease the matrix dimension then I resize it at the original dimension. Now what I am expecting is to see a smaller gaussian profile. But I see this:

The first matrix is the original one, the second one is the resized one. Why we dont see a gaussian.

[Mathieu NOE]

hello

the first image is related to Q or G ? - this is not clear

I ask because you resized G and not Q

[Adam Danz]

What plotting commands are you using to produce each plot?

What's the outputs for

size(G)
size(SS)

[Alberto Paniate]

Sorry, i have edited question. The first image is A, the second one is SS

[Adam Danz]

Ok, but I have the same questions.

[Alberto Paniate]

Take the matrix A:

G=A.*conj(A);
figure,imagesc(G(:,:));

I have to do it because A is a complex matrix.

If I try with size(G) I have this error:

Array indices must be positive integers or logical values.

[Rik]

Don't use size as a variable name.

[Alberto Paniate]

both 1000 1000

[Adam Danz]

Could you provide a mat file containing the variables A and t so we can reproduce this on our end?

[Alberto Paniate]

I have found that is a problem linked with the complexity of the matrix, if we consider a real matrix we dont have this problem.

[Adam Danz]

Thanks but it looks like there's at least 1 variable that's undefined.

Unrecognized function or variable 'E0'.
Error in ProvaSito (line 13)
        A(k,j)=E0*exp(-0.5/(sigma^2)*((k-round(size2/2))^2+(j-round(size2/2))^2)); %gaussiana 
>> 

[Alberto Paniate]

put E0= 10; it is the same sorry

[Adam Danz]

Still more undefined vars.

Unrecognized function or variable 'gran'.
Error in ProvaSito (line 17)
for k=1:gran:(size2-gran) 

Perhaps you could download your attachment and try to run it after clearning your workspace. Then you can find the missing vars and supply them.