Matrix Subtraction by taking mean of same matrix

I want to subtract of data which is of 413 X 264 and from taking a mean of same data with 45x45 matrix

My code is

winSz = [45,45];   % window size [width (x), height (y)]
[m, n] = size(HH_lin_0p2km);
winX0 = 1:winSz(1):n-winSz(1)+1;   % starting index of x-values for each window
winY0 = 1:winSz(2):m-winSz(2)+1;   % starting index of y-values for each window


xWin = 0:winSz(1)-1;
yWin = 0:winSz(2)-1; 
for i = 1:size(HH_C,1)
    for j = 1:size(HH_C,2)
        HH_C = HH_lin_0p2km(winY0(i)+yWin, winX0(j)+xWin)-nanmean(HH_lin_0p2km(winY0(i)+yWin, winX0(j)+xWin),'all');
        HV_C = HV_lin_0p2km(winY0(i)+yWin, winX0(j)+xWin)-nanmean(HV_lin_0p2km(winY0(i)+yWin, winX0(j)+xWin),'all');
    end
end

And I am getting error as:

I am looking for output in 413x264 MATRIX

ndex exceeds the number of array elements. Index must not exceed 5.
Error in Soil_Moisture_Estimation (line 492)
        HH_C = HH_lin_0p2km(winY0(i)+yWin, winX0(j)+xWin)-nanmean(HH_lin_0p2km(winY0(i)+yWin, winX0(j)+xWin),'all');

Any one please help

Answer 1

[Matt J]

It would be faster to use sepblockfun from the File Exchange,

https://www.mathworks.com/matlabcentral/fileexchange/48089-separable-block-wise-operations

HH_C = process(HH_lin_0p2km);
HV_C = process(HV_lin_0p2km);


function out=process(H)

 winSz = [45,45];   % window size [width (x), height (y)]
 [m, n] = size(H);
 mm=ceil(winSz(1)/m)*m;
 nn=ceil(winSz(2)/n)*n;

 if mm*nn>m*n, H(mm,nn)=0; end
 
 nanmap=isnan(H);
 H(nanmap)=0;
 
 Means=sepblockfun(H,winSz,'sum')./sepblockfun(nanmap,winSz,'sum');
 
 out=H-repelem(Means,winSz); 
 
 out(nanmap)=nan;
 out=out(1:m,1:n);

end

[Preet Lal]

What's process here?

It's not working

[Matt J]

How about the following modified version? I've run it on a 2x2 winSz as a test.

winSz=[2,2];

H=randi(4,4)

H = 4×4

3 2 4 3 1 3 2 1 3 4 1 3 4 3 3 3

[H_C,blockMeans] = process(H,winSz)

H_C = 4×4

0.7500 -0.2500 1.5000 0.5000 -1.2500 0.7500 -0.5000 -1.5000 -0.5000 0.5000 -1.5000 0.5000 0.5000 -0.5000 0.5000 0.5000

blockMeans = 4×4

2.2500 2.2500 2.5000 2.5000 2.2500 2.2500 2.5000 2.5000 3.5000 3.5000 2.5000 2.5000 3.5000 3.5000 2.5000 2.5000

function [out,Means]=process(H,winSz)

 [m, n] = size(H);
 mm=ceil(winSz(1)/m)*m;
 nn=ceil(winSz(2)/n)*n;

 if mm*nn>m*n, H(mm,nn)=0; end
 
 nanmap=isnan(H);
 H(nanmap)=0;
 
 Means=sepblockfun(H,winSz,'sum')./sepblockfun(~nanmap,winSz,'sum');
 Means=repelem(Means,winSz(1),winSz(2));
 
 
 out=H-Means; 
 
 out(nanmap)=nan;
 out=out(1:m,1:n);

end