3D Curve Fitting

If your desired equation (z=f(x,y)) is a linear function, then you can use the LINEST() function to obtain the parameters of the function. (Note that "linear" here is more of a linear algebra sense, so that it can be used for much more than straight lines and planes.) https://support.office.com/en-us/art...a-fa7abf772b6d

If your desired equation is not linear and cannot be transformed to something linear, then you can use non-linear regression techniques that involve the Solver utility to optimize your desired objective function. https://support.office.com/en-us/art...b-f63e45925040

What kind of regression are you trying to do?

With name like "Spline", I would guess it is using some kind of cubic spline algorithm. How important is it to your algorithm that you exactly match what this previous "spline.dll" did?

If it is necessary to exactly match what spline.dll did, then it will likely be necessary to know what spline.dll was doing.

With four points, I am not sure what difference there would be between a simple cubic polynomial regression and a cubic spline. If you are not terribly concerned with exactly matching spline.dll, then a first guess might be to try a simple cubic polynomial regression using LINEST() (see help file linked above).

Beyond that, I'm not sure what to recommend without knowing something more about the desired regression equation.