Sam Brilleman

> Thanks @sambrilleman for setting this up. At this stage I am trying to make sense of your code / approach. So some questions follow every now and then. Once...

@bgoodri @jgabry I'm hoping you can maybe help me with some methodology and implementation. For `stan_surv` I am formulating a regression model defined on the hazard scale, that allows for...

@bgoodri Thanks, that makes sense. But just to elaborate, suppose I wish to specify the coefficients with MVN prior with penalty matrix, is it possible to specify a non-centered parameterisation...

@ermeel86 Thanks, I would love for your advice on this! I had tried to implement the prior distribution as proposed in Lang and Brezger's "[Bayesian P-splines](https://www.jstor.org/stable/1391151)" paper, but got completely...

@csetraynor Thanks, I think I've pretty much sorted a `plot.stansurv` method for the baseline hazard and the time-dependent hazard ratio, and otherwise the method calls `NextMethod` (i.e. `plot.stanreg`) which seems...

@ermeel86 @bgoodri Just to follow up on this. I've opened a pull request with a first working version of `stan_surv` that I hope is ready for release. I've just implemented...

Hi Karl I've not yet implemented hierarchical parameters for `stan_surv`. Sorry that the error message was so useless, I will fix that. Hierarchical survival models will happen eventually. But I...

@karlsmithbyrne If you want to go the Bayesian route with a proportional hazards model, then I'm not sure there are many alternatives. I think **brms** currently fits exponential and Weibull...

Hey @padpadpadpad... so sorry it has taken me this long to get back to you!! > I would recommend possibly putting some code up on how to install the specific...

> I did manage to get the code working though so that's great. In ps3, we are predicting for a new level of the random effect and marginalising over the...