Time to Event Model with Smooth Dose-Response Curve

This R walkthrough guides through data generation, prior calibration and posterior simulation for a piecewise exponential model with a NDLM for the underlying dose-response curve.

Highlights from this walkthrough include:

• Review of survival analysis
• NDLM for smooth curves
• How to generate data from a piecewise exponential model
• Prior calibration for a piecewise exponential model
• Efficient model fitting

Note: the piecewise exponential model can be fitted as a Poisson regression with a reparametrization trick. However, in practice I have observed that the model fitting procedure proposed here (using the likelihood explicitly) is not only faster but is also mixing better when run with the same number of iterations.