/robjm

Robust Joint Models

Primary LanguageRebol

robjm: An R Package for Robust Longitudinal and Joint Models

Dynamic predictions of kidney graft survival in the presence of longitudinal outliers

Robust joint models

Install and load using the following lines:

library(devtools)
install_github("ozgurasarstat/robjm")
library(robjm)

TO DO:

  • add rate of change to the survival model other than nor nor and mod3

  • prediction for the models other than nor and mod3

  • prediction for longitudinal data

  • plot, summary, print etc

  • develop general function for ld and jm, e.g. a function that is for nor, mod1, mod2, mod3, tv as a sinlge file

  • and make it or all functions efficient, eta1 and eta2, linpred removed, for loop for y is removed, for loop for B or Bstar is removed. loop for d_B etc is changed

  • add last.time option

  • looping over A[n] where A is an array is more fficienct than if A is a matrix (pp 325 of stan ref doc)

  • add information criterion for number of knots

  • make the codes more efficients