Cameron Davidson-Pilon

The first three questions can be answered by checking out the docs: https://lifelines.readthedocs.io/en/latest/fitters/regression/WeibullAFTFitter.html?highlight=plot_partial_effects_on_outcome#lifelines.fitters.weibull_aft_fitter.WeibullAFTFitter.plot_partial_effects_on_outcome I don't quite understand your 4th question, however

> Is there a confidence interval band fill between for the survival plot? If not, how can I get it? Unfortunately, not

cumulative hazard (t) = int_0^t hazard(s) ds so it's possible for cumulative_hazard < hazard. Think about how a short spike in a function might affect its integral.

Sure, yea, but you should expect very small variation in your log10 variable as a consequence.

Wow that's very impressive! One thing I think you should try is to bin the times into buckets (as tied times are common in survival datasets, as we are often...

Very exciting work, team!

I'm impressed. I'm going to have to try this library locally. Is the following (mostly) correct? One significant speed up is from using an approximation to the Hessian. This approximation...

Hi @bayesfactor! Thanks for the PR (sorry about the delay). We have an `lifelines.utils.restricted_mean_survival_time` function now - have you compared against that? We don't have a `difference_` though, so that...

Hi @bayesfactor, Thanks for keeping up with this! Can you explain how this works for parametric fitters? In my head, to compute the AUC of a parametric fitter, some `scipy.integrate.quad`...

I think we can combine the functions! Have a global `restricted_mean_survival_time` that, based on the fitter, chooses an implementation. The [current implementation for computing the variance for KMF's](https://github.com/CamDavidsonPilon/lifelines/blob/master/lifelines/utils/__init__.py#L317-L320) is imprecise...