survModelData: Process the data needed for survival models

In hypergsplines: Bayesian model selection with penalised splines and hyper-g prior

Description

Process the data needed for survival models

Usage

  survModelData(times, X, observed,
    continuous = rep.int(TRUE, ncol(X)), nKnots = 4L,
    splineType = "linear",
    gPrior = HypergnPrior(a = 4, n = sum(observed)))

Arguments

times	the numeric vector of survival times
X	the numeric matrix of covariates (not including time)
observed	the logical vector of observation indicators, TRUE entries represent truly observed survival times, FALSE entries represent censored survival times.
continuous	see glmModelData for details
nKnots	see glmModelData for details
splineType	see glmModelData for details
gPrior	see glmModelData for details. Defaults to the hyper-g/n prior where n is chosen as the number of events, instead of the number of observations.

Value

a list with the internally needed results.

Author(s)

Daniel Sabanes Bove [e-mail]

Examples

## get some data
library(survival)
pbc <- na.omit(pbc)
pbc$sex <- as.numeric(pbc$sex == "f")

## try the function
md <- survModelData(times=pbc$time,
                    X=
                    as.matrix(subset(pbc,
                                     select=
                                     c(trt,
                                       age,
                                       sex,
                                       ascites,
                                       hepato,
                                       spiders,
                                       edema,
                                       bili,
                                       chol,
                                       albumin,
                                       copper,
                                       alk.phos,
                                       ast,
                                       trig,
                                       platelet,
                                       protime,
                                       stage))),
                    observed=
                    pbc$status == 2,
                    continuous=
                    c(FALSE,
                      TRUE,
                      FALSE,
                      FALSE,
                      FALSE,
                      FALSE,
                      FALSE,
                      TRUE,
                      TRUE,
                      TRUE,
                      TRUE,
                      TRUE,
                      TRUE,
                      TRUE,
                      TRUE,
                      TRUE,
                      FALSE),
                    nKnots=6L,
                    splineType="cubic",
                    gPrior="hyper-g/n")

## look at the results
str(md)

hypergsplines documentation built on May 2, 2019, 6:14 p.m.