Code and data sets for the
methods discussed in Jeffrey S. Simonoff , "Three Sides of Smoothing:
Categorical Data Smoothing, Nonparametric Regression, and Density Estimation,"
*International Statistical Review*, **66**, 137-156 (1998).
After downloading the S-PLUS dump file three.dmp, restore it using the
data.restore command (the `foreign` package must be loaded first if R is being used). The following functions and data sets will be restored:

aicc.loess
aicc.spline
cat.crit.loess
cat.crit.loess2
cat.crit.spline
crit.loess
crit.spline
loclike.de
loclike2d.de
penlike.de
bondrat.dat
college.dat
female.dat
lottery.dat
readscore.dat
stock.dat
undergrad.dat (Note: Table 4 in the paper has a typographical error. The entry
under (Importance of size of salary = 4, Importance of employment opportunities =
4, Chose major as a high school student) should be 8, not 0. The local
quadratic probability estimates in Table 5 are correct.)

The file is available here.

Click here for related code for nonparametric
regression and semiparametric
and additive models.

Click
to go to my homepage