smoothSurv (version 1.6)

eval.Gspline: Evaluate a G-spline in a grid of values

Description

This function computes values of $$f(x) = \sum_{j=1}^g c_j \varphi_{\mu_j, \sigma_j^2}(x)$$ in a grid of $x$ values.

In above expression, $phi(x; mu[j],sigma[j]^2)$ denotes a density of $N(mu[j], sigma[j]^2)$.

Usage

eval.Gspline(Gspline, grid)

Arguments

Gspline

A data frame with at least three columns named ``Knot'', ``SD basis'' and ``c coef.'' which determine $mu[1], ..., mu[g]$, $sigma[1], ..., sigma[g]$ and $c[1], ..., c[g]$. Data.frame with such properties can be found e.g. as spline component of the resulting object returned by functions smoothSurvReg and minPenalty.

grid

A numeric vector giving the grid of $x$ values at which the G-spline is to be evaluated.

Value

values).

Examples

  spline <- minPenalty(knots=seq(-4.2, 4.2, by=0.3), sdspline=0.2, difforder=3)$spline
  values <- eval.Gspline(spline, seq(-4.5, 4.5, by=0.05))
  plot(values, type="l", bty="n", lwd=3)