Skip to content

The flexsurv R package for flexible parametric survival and multi-state modelling

Notifications You must be signed in to change notification settings

chjackson/flexsurv

Repository files navigation

flexsurv

The flexsurv R package for parametric survival and multi-state modelling.

Key features

  • Parametric models for time-to-event (survival) data. Data may be right-censored, and/or left-censored, and/or left-truncated. Several built-in parametric distributions are available, including a very flexible model based on splines (Royston and Parmar).

  • Any user-defined parametric model can be employed by supplying R functions defining the distribution.

  • Covariates can be included using a (log-)linear model on any parameter of any distribution. This typically defines an accelerated failure time or proportional hazards model.

  • Multi-state models for continuously-observed data can be defined by piecing together transition-specific parametric models of any kind. (For intermittently-observed data, see instead the msm package.)

Learn more

User guide (PDF).

Guide to multi-state modelling in flexsurv.

Full reference manual for all the package's functions.

Paper in Journal of Statistical Software (Jackson 2016). Mostly the same as the user guide, but not kept up to date.

Installation (stable CRAN version)

install.packages("flexsurv")

Installation (development version)

install.packages("devtools") # if devtools not already installed
devtools::install_github('chjackson/flexsurv-dev')

Lifecycle: stable CRAN status R-CMD-check Codecov test coverage