ICARBHSampler {BayesPiecewiseICAR} | R Documentation |
This function fits a piecewise hazard using a hierarchical model with a ICAR dependence applied to the hazard heights
ICARBHSampler(Y, I, B, hyper)
Y |
This is a n-vector containing patient survival times |
I |
This is a n-vector containing patient censoring indicators (0 for censored patient) |
B |
Number of iterations to run the sample |
hyper |
Vector of hyperparameters. In order, this contains a1, b1 which are the inverse gamma hyperparameters on sigma^2. Phi which is the hyperparameter on the mean number of split points. Jmax which is the maximum allowed number of split points. cl1 which is a tuning parameter greater than 0. J1 is the starting number of split points for the MCMC. Finally, clam1 which is between 0 and 1 and characterizes the spatial dependency of the baseline hazard heights. |
Returns a list containing the posterior samples of the split points, split point locations, log hazard rates and hierarchical samples
https://adventuresinstatistics.wordpress.com/2016/07/29/bayespiecewiseicar-tutorial-and-details/
Lee, K. H., Haneuse, S., Schrag, D. and Dominici, F. (2015), Bayesian semiparametric analysis of semicompeting risks data: investigating hospital readmission after a pancreatic cancer diagnosis. Journal of the Royal Statistical Society: Series C (Applied Statistics), 64: 253-273.
####This generates random survival data Y=rexp(100,1/20) I=rbinom(100,1,.5) ###Sets hyperparameters a1=.7 b1=.7 phi=3 Jmax=20 cl1=.25 clam1=.5 J1=3 ###Combines the hyperparameters in to a vector hyper=c(a1,b1,phi,Jmax,cl1,J1,clam1) ###Set Number of iterations B=100 ###Run the Sampler X=ICARBHSampler(Y,I,B,hyper) X