Reusing natural neighbor interpolation

I have a problem where I need to interpolate data onto a relatively large mesh for many thousands of time steps, and the data location does not change with time step. Is there a way to extract the natural neighbor weightings in order to perform a simple matrix multiplication of the dataset against the mesh coordinates rather than performing the full interpolation routine at each time step?

[Alessandro]

That's the exact thing I need as well. I tried stepping into the function but I couldn't find a way to store the weights vector. Did you find a solution for this yet?

Answer 1

[KSSV]

You can select an area (sub matrix from the matrix) around the point of interpolation and proceed as you were proceeding with original matrix.