convert a x y z data(1D) to MxN matrix without having NaN values in Z matrix?

I have a XYZ sattered data all are having same number of elements and are 1D. While interpolating, I get MxN matrix for X Y Z; but Z has some NaN value outside the scattered data boundary. I don't want to violet the surface, what to do?

[KSSV]

Show us your data and code. If the data lies outside, it comes under extrapolation and we cannot trust the extrapolated result.

Answer 1

[the cyclist]

If you are using interp3 (or related functions), extrapolated values will automatically be set to NaN, for some interpolation methods (e.g. the default linear interpolation). This makes sense for methods that rely on calculating values in between two endpoints.

See that documentation link for methods that will allow for non-NaN extrapolation.