About Data Fitting Methods

If you are creating a contour graph or surface graph in the Design Gateway or the Runtime Gateway, you can determine the data fitting method that Isight uses to create a three-dimensional surface from a set of data points.

You can define the default data fitting method (Simple Fit, RSM Approximation, or RBF Approximation) using the Isight preferences (see Setting Postprocessing Preferences).

This page discusses:

Simple Fit Technique

The Simple Fit technique generates a surface that passes through all of the actual data points. Simple Fit is the fastest and least complex technique. However, it can be inaccurate in regions where no data exist, particularly on the edges of the regions being viewed. This data fitting method is the default.

RSM Approximation Technique

The RSM Approximation technique automatically creates a Response Surface Model (RSM) approximation with quadratic polynomial order for the selected data points. Isight creates a surface graph of this quadratic approximation. You can use this option to smooth out the noise in the data. If the range of inputs is large and/or the shape of the underlying data is very complicated, significant differences can result between the underlying data points and the displayed surface graph. This option is more complex and time-consuming than the Simple Fit method.

RBF Approximation Technique

The RBF Approximation technique automatically creates a Radial Basis Function (RBF) approximation for the selected data points. Isight creates a surface graph of this RBF approximation. You can use this option to smooth out the noise in the data. RBF approximation is capable of following complicated data more closely than the RSM approximation, which reduces the potential differences between the underlying data points and the displayed surface graph. This option is the most complex and time-consuming and can slow down the responsiveness of the Runtime Gateway while the graph is created.