Why does the GRIDDATA function from MATLAB produce different results than that of a very simple 1-nearest neighbor algorithm?

Why does the GRIDDATA function from MATLAB produce different results than that of a very simple 1-nearest neighbor algorithm?

When I use the nearest neighbor algorithm from GRIDDATA, the results obtained are different from a very simple implemenation, such as:

    xb = 2342323; % x-coordinate of point to search for

    yb = 13254; % y-coordinate of point to search for

    for ii=1:length(X)

      xa=X(ii);

      ya=Y(ii);

      dis=sqrt((xa-xb)^2+(ya-yb)^2);

      if dis<dism

          ind=ii;

          dism=dis;

        end

    end

    zb = Z(ind) % interpolated z-value of searched point

Furthermore, I receive a warning for duplicate points that is not true:

   Warning: Duplicate x-y data points detected: using average of the z values.

   (Type "warning off MATLAB:griddata:DuplicateDataPoints" to suppress this warning.)

   > In D:\MATLAB6p5p1\toolbox\matlab\polyfun\griddata.m at line 59

      In q:\1400113\ex.m at line 24

Why does this occur?

Answer 1

[MathWorks Support Team]

This bug has been fixed in Release 14 Service Pack 2 (R14SP2). For previous releases, please read below for any possible workarounds:

This bug has been fixed for Release 14 SP1 (R14SP1). For previous releases, please read below for any possible workarounds:

GRIDDATA uses a Delaunay triangulation of the data to perform its analysis. When the data has strange scaling, this algorithm might detect points that are very close to each other (relative to the absolute value) and see them as duplicates.

GRIDDATA then uses an averaged z-value for these points, which can lead to different results compared to the simple algorithm.

As a workaround, remove large offsets from the data, so that the ratio of the variance of the data to the mean value of the data increases. The following code illustrates how you can do this:

      % get offset values

      xmin = min(X); 

      ymin = min(Y); 

      % correct for offset 

      X = Y - xmin; 

      Y = Y - ymin;

      %remove offset also fromcoordinates of point to search for

      xb = xb - xmin;  

      yb = yb - ymin;