next up previous
Next: Some Special Spline Models Up: General Smoothing Spline Regression Previous: Model and Estimation

The ssr Function

The S function for fitting a SSR model is ssr. A typical call is

    ssr(formula, rk, data)
Only the first two arguments, formula and rk, are required. Together they describe the response variable and the model space. formula, a statistical formula as in lm, lists the response variable on the left hand side and the bases $\phi_1(t), \cdots, \phi_M(t)$ of ${\cal H}_0$ on the right hand side of an operator ~. rk is an expression that specifies the reproducing kernel $R_1$. Expressions of rk for commonly used reproducing kernels are available in our library. These expressions will be discussed in the following sections. Users can easily add their own expressions of reproducing kernels.

The optional argument data specifies the data frame. Other optional arguments will be discussed in the following sections.

ssr returns an object of class ``ssr'' with many variables representing the fit to the specified smoothing spline model. Detailed information about this object can be found in the help file on ssr.object.

Yuedong Wang 2004-05-19