Constrained Surface of Best Fit
For a smoothed surface of best fit the procedure is the same as for creating a plane of best fit. You first need to setup for each section the grading parameters. If you look at the picture below :
The appropriate parameters to consider are highlighted

Bearing  This defines in which direction the X and Y grades are defined. It does not need to be inline with the defined grid.

Smooth  This needs to be selected.

minX  the software attempts to make sure that no grades are less than this value.

maxX  the software attempts to make sure that no grades greater than this value are created.

minY and maxY  same except along the Y axis.
You do not need to set any of these values. However if you set one of the values, for example min then make sure you set a max value.
New click on the Grading
As per the "plane of best fit" you can do the grading as either based on a grid or the more rigorous based triangle solution.
The following dialog is shown.
It contains a number of parameters that can be adjusted.
Grid Size can be adjusted. We suggest you start with a larger value and decrease the value if you want to refine the design.
The Sampling value is a distance on the surface. The algorithm uses all the points found within this distance when deciding how the surface is to be graded.
The smoothing value refers to a distance and affects how the points are used. Setting a value less than the sampling value means that points towards the edge of the sampling distance are weighted less than those closer to the sampling position.
If you are not sure we suggest using smaller values to see the effect and slowly increase the values until a more optimal solution is achieved.
Please see the tutorials described towards the end of this help.