How can I interpolate a data set to fill a missing point in a vector using MATLAB?

How can I interpolate a data set to fill a missing point in a vector using MATLAB?

If I have values for a function on many of the data points on a grid and want to fill in the values of that function for the missing data points on the grid, what function should I use, and how should I set this up?

Answer 1

[MathWorks Support Team]

The following code evaluates the sample function

 f(x,y,z) = x+2*y+4*z

on most of the integer-valued grid points on the unit cube, and uses those function values to interpolate the function value on an integer-valued grid point for which we do not have the function value.

% Generate data for the unit cube

[xx,yy,zz]=ndgrid(-1:1,-1:1,-1:1);

% Combine the three individual coordinates into a matrix

R=[xx(:) yy(:) zz(:)];

% Remove the entry for (1, -1, 1)

R(23,:)=[];

% Calculate a function on the given points

v = xx.*yy+2*xx.*zz+4*yy.*zz;

% Remove the point corresponding to (1, -1, 1)

v = v(:);

v(23, :) = [];

% Interpolate

% The duplicate entry as the third argument is due to

% a bug in griddatan with respect to row vectors as third

% input arguments

s = griddatan(R, v, [1 -1 1;1 -1 1])