Problem in using scatteredInterpolant

Hi,

I am doing interpolation here to get values from variable z according to the respective lat lon. My variables from hdf file are:

lat = 406x270;

lon = 406x270;

z = 2030x1354;

lat_i = imresize(lat,[2030 1354],'bilinear');
lon_i = imresize(lon,[2030 1354],'bilinear'); 

There's another text file that contains lat lon values in the form of column vector:

lat_txt = tab.lat; % extract the column of latitude values from text file (3798x1 double)
lon_txt = tab.lon; % extract the column of longitude values from text file (3798x1 double)

I've used scatteredInterpolant here for interpolation

F = scatteredInterpolant(lat_i(:), lon_i(:), z(:));
Res = F(lat_txt, lon_txt); % To get values of z according to the lat lon of text file  

Problem is z conatin integers (0,1,2,3,4) and using scatteredInterpolant gives values (1, 1.838, 2.84, 2.139, 2, 2.0022, 1.0014, 3.339, 3.000, 2.779,3 etc). And this thing is effecting my results. Is there a way that I only get in the form of integers (0,1,2,3,4)?

Warning:

 Warning: Duplicate data points have been detected and removed - corresponding values have been averaged.

Secondly, I want 0 and 1 to be removed from z before passing it through scatteredInterpolant. I assign NaN to pixels containing 0 and 1. Not sure whether it's fine or not. Any help will be highly appreciated.

[Walter Roberson]

Problem is z conatin integers (0,1,2,3,4) and using scatteredInterpolant gives values (1, 1.838, 2.84, 2.139, 2, 2.0022, 1.0014, 3.339, 3.000, 2.779,3 etc). And this thing is effecting my results. Is there a way that I only get in the form of integers (0,1,2,3,4)?

No

[Zhou Ci]

@Walter Roberson what can possibly be done to avoid getting such values (1, 1.838, 2.84, 2.139, 2, 2.0022, 1.0014, 3.339, 3.000, 2.779,3). Any other function in MATLAB other than scatteredInterpolant?

[Walter Roberson]

Nothing can be done for that purpose. If you need integers then you should not be doing any kind of interpolation.

Answer 1

[Bruno Luong]

Use nearest method

F = scatteredInterpolant(..., 'nearest')