(View the complete code for this example.)
This example illustrates several types of spline fit functions that are available in ODS Graphics and SAS/STAT. It begins by showing you how to use PROC SGPLOT to fit a single function through a scatter plot. Later examples use PROC SGPLOT to fit multiple functions by specifying a group variable. Subsequent sections show how to fit spline functions in SAS/STAT procedures, concentrating on PROC TRANSREG and the spline functionality in the EFFECT and EFFECTPLOT statements. PROC ADAPTIVEREG is also discussed. Topics include polynomials, polynomial splines (Smith 1979), B-splines (De Boor 1978), penalized B-splines (Eilers and Marx 1996), monotone splines (Winsberg and Ramsay 1980), natural cubic splines and restricted splines (Hastie, Tibshirani, and Friedman 2009), smoothing splines (Reinsch 1967), multivariate adaptive regression splines (Friedman 1991), and loess (Cleveland, Devlin, and Grosse 1988).