Cameron Davidson-Pilon

icon indicating a website link https://dataorigami.net icon indicating an email [email protected]

icon company @Pioreactor icon location Waterloo, Canada icon location CEO of Pioreactor. Former Director of Data Science @Shopify. Author of Bayesian Methods for Hackers and DataOrigami.

Results 297 comments of Cameron Davidson-Pilon

CoxPHFitter has no attribute '_preprocess_dataframe'

Hi @isirisiss, are you able to provide a full stack trace?

CoxPHFitter has no attribute '_preprocess_dataframe'

Oh, it's because you're using `spline` - that's not implemented, my bad. I don't know a code work-around for now, unfortunately, but you could check the assumptions using a non-spline...

Using weighted data in proportional_hazard_test() for CoxPH

Thanks for the detailed issue @aongus, I'll look into this asap

Using weighted data in proportional_hazard_test() for CoxPH

ack sorry, it's a high priority but am stuck on it. I haven't made much progress, unfortunately.

Using weighted data in proportional_hazard_test() for CoxPH

Hi @aongus, I've dug a bit into this recently, and the problem may be due to R changing their algorithm recently for computing these values, see https://github.com/CamDavidsonPilon/lifelines/issues/997#issuecomment-652567848

Set reference level in CPH using formula with Formulaic?

Hi @t-silvers, AFAIK formulaic doesn't support that feature yet. If you really need that feature, you can use Patsy to transform your df _outside_ lifelines, and then pass the transformed...

if the first entry was after 0, this min would produce a nan

why didn't I merge this?

kmf.fit -> is_left_censoring?

I actually apologize - previous Cameron _way over-engineered_ this class and made us all confused today 🙃 Anyways, I think the _correct_ behaviour is as follows: - `plot` should probably...

baseline functions differ in R and Python

Hi @Valery2511 At first thought, I'm not sure. I do have tests in the project comparing [lifelines baseline vs R's baseline](https://github.com/CamDavidsonPilon/lifelines/blob/master/lifelines/tests/test_estimation.py#L5140). I will have to do some digging. Are you...

baseline functions differ in R and Python

Hi @PortlandMichelle. Hard to say what might be going on without seeing the dataset used. Can you reproduce it with a small synthetic dataset?