How can I average every other value in a 500 by 500 matrix

I have a matrix that is missing every other value and i am trying to interpolate it by taking the value of the number before and after and averaging the 2

this is what I have so far but it is not working any help is greaty appreciated!

[M,N] = size(B)

for i = 1:N
    if N(i) == 0
        H(i) = (N(i-1) + N(i+1))/2
    else
        H(i) = N(i)
    end
end

for j = 1:M
    if M(j) == 0
        K(j) = (M(j-1) + M(j+1))/2
    else
        K(j) = M(j)
    end 
end
disp(B)

[Walter Roberson]

Is the pattern

0 A 0 B 0 C
0 D 0 E 0 F
0 G 0 H 0 I

or is it

0 A 0 B 0 C
D 0 E 0 F 0
0 G 0 H 0 I

or is it

0 A 0 B 0 C
0 0 0 0 0 0
0 G 0 H 0 I

[Martin Jordan]

the pattern is

0 A 0 B 0 C

D 0 E 0 F 0

0 G 0 H 0 I

[per isakson]

See

https://se.mathworks.com/matlabcentral/answers/397278-interpolation-of-2d-matrix-using-interp2-to-eliminate-nan#answer_317124

Answer 1

[Kritika Bansal]

Hi,

There are few ways to do so in MATLAB. Some of them are listed below:

1. Using interp2 function of MATLAB. As per my understanding of this function, if your data is non gridded, then there is a possibility that all the missing values are not interpolated. You can read more about this function here: https://www.mathworks.com/help/matlab/ref/interp2.html
2. Using inpaint_nans function from the file exchange available at https://www.mathworks.com/matlabcentral/fileexchange/4551-inpaint_nans

If you want to do the averaging of the neighbours as you mentioned, you can take a look at the following code:

%x is the input matrix
[m,n] = size(x);

y=zeros(m,n);
for i=1:m
    for j=1:n
        if x(i,j)==0
            if j==1
                l=j;
            else
                l=j-1;
            end
            if j==n
                r=j;
            else
                r=j+1;
            end
            val = (x(i,l)+x(i,r))/2;
            y(i,j) = val;
        else
            y(i,j) = x(i,j);
        end
    end
end