Paul Lambert
  • Home
  • Publications
  • Software & Tutorials
  • Talks
  • Courses
  • Interactive graphs

On this page

  • Fitting stpm3 models.
  • Predictions (conditional on covariate values)
  • Marginal predictions
  • Other Details
  • Releases

stpm3

stpm3 fits flexible parametric survival models. These models use splines to model the effect of the time scale. stpm3 replaces stpm2, though stpm2 will continue to be available on SSC.

I have added some examples to explain some of the features of stpm3 and differences to stpm2.

Installing You will need to install the stpm3, standsurv and gensplines packages. You can install from within Stata from SSC using the code below. If you have already installed earler test versions from my website then I recommend first removing these versions using ado uninstall before installing the version from SSC. Stata does not always refresh the mata libraries when using ado update, so I suggest restarting Stata after updating.

ssc install stpm3
ssc install standsurv
ssc install gensplines

Fitting stpm3 models.

  • Change to the scale option
  • Use of factor variables

Predictions (conditional on covariate values)

  • Frames for prediction
  • Extended functions
  • Multiple at options and contrasts
  • Relative survival models

Marginal predictions

  • Using standsurv with factor variables

Other Details

  • [Numerical integration for stpm3 models on log hazard scale]

Releases

See stpm3_releases.txt