predictProb.survfit

Extracts predicted survival probabilities for survival models fitted by <code>survfit</code>, providing an interface as required by <code>pmpec</code>.

models

regression

survival

Designed for prediction error estimation
through resampling techniques, possibly accelerated by parallel
execution on a compute cluster. Newly developed model fitting
routines can be easily incorporated. Methods used in the package are detailed in
Porzelius Ch., Binder H. and Schumacher M. (2009) <doi:10.1093/bioinformatics/btp062>
and were used, for instance, in
Porzelius Ch., Schumacher M.and Binder H. (2011) <doi:10.1007/s00180-011-0236-6>.

Frederic Bertrand

peperr

Parallelised Estimation of Prediction Error

predictProb.survfit function

<dl><dt>object</dt>
<dd>a fitted model of class <code>survfit</code>.</dd><dt>response</dt>
<dd>survival object (with <code>Surv(time, status)</code>, where time is an <code>n</code>-vector of censored survival times and status an <code>n</code>-vector containing survival status, coded with 0 and 1.</dd><dt>x</dt>
<dd><code>n*p</code> matrix of covariates.</dd><dt>times</dt>
<dd>vector of evaluation time points.</dd><dt>train.data</dt>
<dd>not used.</dd><dt>...</dt>
<dd>additional arguments, currently not used.</dd></dl>

Arguments

Extract predicted survival probabilities from a survfit object — predictProb.survfit

<dl>

<dt>object</dt>
<dd>a fitted model of class <code>survfit</code>.</dd>

<dt>response</dt>
<dd>survival object (with <code>Surv(time, status)</code>, where time is an <code>n</code>-vector of censored survival times and status an <code>n</code>-vector containing survival status, coded with 0 and 1.</dd>

<dt>x</dt>
<dd><code>n*p</code> matrix of covariates.</dd>

<dt>times</dt>
<dd>vector of evaluation time points.</dd>

<dt>train.data</dt>
<dd>not used.</dd>

<dt>...</dt>
<dd>additional arguments, currently not used.</dd>

</dl>

predictProb.survfit: Extract predicted survival probabilities from a survfit object

Description

Usage

Value

Arguments