smooth.cons {survPen}

Design and penalty matrices of penalized splines in a smooth.spec object

Description

Builds the design and penalty matrices from the result of smooth.spec.

Usage

smooth.cons(
  term,
  knots,
  df,
  by = NULL,
  option,
  data.spec,
  same.rho = FALSE,
  name
)

Arguments

term

Vector of strings that generally comes from the value "term" of a smooth.spec object.

knots

List of numeric vectors that specifies the knots of the splines (including boundaries).

df

Degrees of freedom: numeric vector that indicates the number of knots desired for each covariate.

by

numeric or factor variable in order to define a varying coefficient smooth; default is NULL.

option

"smf", "tensor" or "tint".

data.spec

data frame that represents the environment from which the covariate values and knots are to be calculated; default is NULL.

same.rho

if there is a factor by variable, should the smoothing parameters be the same for all levels; default is FALSE.

name

simplified name of the smooth.spec call.

Value

List of objects with the following items:

X

Design matrix

pen

List of penalty matrices

term

Vector of strings giving the names of each covariate

knots

list of numeric vectors that specifies the knots for each covariate

dim

Number of covariates

all.df

Numeric vector giving the number of knots associated with each covariate

sum.df

Sum of all.df

Z.smf

List of matrices that represents the sum-to-zero constraint to apply for "smf" splines

Z.tensor

List of matrices that represents the sum-to-zero constraint to apply for "tensor" splines

Z.tint

List of matrices that represents the sum-to-zero constraint to apply for "tint" splines

lambda.name

name of the smoothing parameters

Examples

library(survPen)

# standard spline of time with 4 knots (so we get a design matrix with 3 columns 
# because of centering constraint)

data <- data.frame(time=seq(0,5,length=100))
smooth.c <- smooth.cons("time",knots=list(c(0,1,3,5)),df=4,option="smf",
data.spec=data,name="smf(time)")