removing boxes with a low amount of data points?

Hello,

I have scattered data points on a 3127 x 254 grid. I would like to interpolate this onto a regular grid (which I have done with triscatteredInterp) but now I would like to find and remove grid boxes that contained less than 30 data points when interpolation took place.

Does anyone know how I could do this?

thanks, Michael

[Michael]

anybody have any ideas?

[Geoff Hayes]

Michael - can you provide a short example of the code you have used to interpolate the data onto the grid? Perhaps that will give others a starting point from which they can attempt to help you with your problem.

[Michael]

Here is my code:

[x,y]=meshgrid(0:2:360,-66:2:66);

x=x(1:end-1,1:end-1); y=y(1:end-1,1:end-1);

yy = y(:,1);
xx = x(1,:);

a = areaquad(yy-(2/2),yy*0,yy+(2/2),(yy*0)+2);
repa = repmat(a,1,180);

mask = zeros(size(depth)); 
mask(find(depth > 0)) = 1;
mask = permute(mask,[2 1]);

F=griddedInterpolant({latt,lonn'},mask);
m=F(y,x);
m(find(m<1 & m>0)) = 1;

repam = repa.*m;
repam(find(repa ==0)) = NaN;
repam = repmat(repam,[1 1 663]);
repam = permute(repam,[3 1 2]);

AM = zeros(663,66,180);

for t = 1:663;
    t
    tic
    a=squeeze(s(t,:,:));

      a=a(:);

      bad1=find(isnan(a)==0);
      a1  = a(bad1);

      lon=REF_lon(:); 
      lon1 = lon(bad1);
      lat=REF_lat(:);
      lat1 = lat(bad1); 

      F1=TriScatteredInterp(lon,lat,a,'natural');

      a_m=F1(x,y);
      AM(t,:,:)=a_m;

    toc

end

[Image Analyst]

Note: requires Mapping Toolbox - wish I'd know that before I wasted my time. I added it to the products list below so others don't waste theirs.

[Michael]

Sorry Image Analyst, I was unaware myself that I was using the mapping toolbox.

Answer 1

[Image Analyst]

Since I can't try your code, all I can suggest is to do a convolution on your array, if you can get your array that is basically a "map" with counts as the values:

window = ones(7);  % Look everywhere in a sliding 7 by 7 window
% Count in each window
sumInWindows = conv(theCountArray, window, 'same');
% Find where count is below, say, 30
binaryImage = sumInWindows  < 30;
image(binaryImage);