\documentclass{article}[11pt]
\usepackage{Sweave}
\usepackage{amsmath}
\addtolength{\textwidth}{1in}
\addtolength{\oddsidemargin}{-.5in}
\setlength{\evensidemargin}{\oddsidemargin}
%\VignetteIndexEntry{Other vignettes}

\title{Other Vignettes}
\author{Terry Therneau}

\begin{document}
\maketitle

A parallel source for the survival package is the therneau/survival directory on
github.
This copy contains several extra vignettes, in the vignette2 directory, that
are not part of the CRAN copy.

\begin{enumerate}
   \item tutorial.Rnw: This works through the examples found in the excellent
     tutorial by H. Putter, M. Fiocco and R. Geskus. 
     Tutorial in biostatistics: competing risks and multi-state models,
     Statistics in Medicine, 2006.
   \item sas.Rnw: An expanded discussion answering some repeated queries about 
     how survival package results differ from SAS.
   \item ridge.Rmd: Addresses a recurrent question wrt using the ridge() function
     with a large number of covariates.
   \item royston.Rnw: A small note that describes the data sets used in an
     excellent paper on validation, and how those data were incorporated into
     the survival package. (Royston and Altman, DOI:10.1186/1471-2288-13-33)
   \item tests.Rnw: A harder look at what SAS ``type 3'' tests are.
   \item pseudo.Rnw: Use of pseudovalues and the pseudo function.
   \item singer.Rnw: Worked examples from a texbook by Singer and Willett.
\end{enumerate}

Number 1 is not on CRAN because it makes use of a data set from the mstate 
package.  As a recommended package survival cannot depend on anything outside
of the base + recommended R.  (A minimally sufficient set that is self contained
is useful for R maintainance and validation.) 
Multi-state models can be computed using either mstate or survival, and this
vignettes shows how parallel the two packages are wrt to what is computed,
though with styles in the setup of the data.  It also provides an extended
validation check, for both packages.

Number 2 delves into relative computing time questions, and those examples take
a very long time to run.
Material in the third vigette is now discussed at a higher level in the 
population estimate and adjusted survival curves vignettes.
Number 4 uses packages geepack and survey to compute corrected standard errors;
packages that are outside the recommended list.

In time, we expect this set of other vignettes to grow.  
\end{document}