Computing in STK4080 autumn 2012
We will use the the programming language R for the practical computing in the lectures as well as the exercises in STK4080.
Below will be given:
- information on some additional R packages that will be used in the course
- information on some of the data sets that will be used for illustration in the lectures and the exercises
- R code for the examples and illustrations
The package "mstate" may be used to estimate transition probabilites in competing risks and Markov chain models. A description of the package is given in a paper by de Wreede et al. in the Journal of Statistical Software .
The package "parfm" may be used to fit parametric frailty models. A description of the package is given in a paper by Munda et al. in the Journal of Statistical Software
R code used in the lectures:
Week 37: Commands for computing Nelson-Aalen estimates are given here .
Week 38: Commands for computing Kaplan-Meier estimates are given here .
Week 39: Commands for computing estimates for competing risks (and Markov chains) are given here .
Week 40: Commands for computing non-parametric tests are given here .
Week 41: Commands for estimating regression coefficients, cumulative hazards, and survival functions for Cox regression are given here .
Week 42: Commands for estimation in stratified Cox regresssion and commands for model check are given here .
Week 43: Commands for additive regression are given here .
Week 45: Commands for Poisson regression are given here .
Week 46: Commands for parametric frailty models are given here .
Week 47: Commands for computing empirical Bayes estimates of individual frailties are given here .
R code for trial project
R commands for the trial project are given here .