We model the log-cumulative baseline hazard for the Cox model via Bayesian, monotonic P-splines. This approach permits fast computation, accounting for arbitrary censorship and the inclusion of nonparametric effects. We leverage the computational efficiency to simplify effect interpretation for...
