Curve Fitting Toolbox 2.1

Fitting Models and Methods

Curve Fitting Toolbox provides a variety of fitting methods, including:

• Linear regression
• Nonlinear regression (trust-region, Levenberg-Marquardt, and Gauss-Newton algorithms)
• Interpolation (linear, biharmonic, nearest neighbor, cubic spline, shape-preserving spline)
• Smoothing (moving average, loess, lowess, smoothing spline, and Savitzky-Golay)

You can specify your own custom equations or choose from a library of linear, nonlinear, and nonparametric fitting models, including:

• Polynomials (to ninth degree for curves and fifth degree for surfaces)
• Exponential functions
• Rational (to degree 5/5)
• Peak (Gaussian)
• Distribution (Weibull)
• Fourier and power series
• Sum of sines

You can assign both weights and bounds on your coefficients. You can also choose from two forms of robust fitting: bisquare or least absolute residual.

	The Surface Fitting Tool (left) and the Curve Fitting Tool (below, left). You can evaluate goodness of fit using a combination of descriptive statistics, visual inspection, and validation.