censored
censored copied to clipboard
tidymodels
Parsnip wrappers for survival models
[Edit] Allow a flexible specification of the reference observation, most likely by providing a data frame, and re-evaluate the choice of default for it. --------------- This really belongs as an...
Hello! Following up on #172, `aorsf` has made it to CRAN and its API is stable.1 - I added wrappers for `aorsf` that were based on the `partykit` templates -...
Hi @hfrick, This PR is a start for addressing https://github.com/tidymodels/censored/issues/115. I have marked the PR as a draft because I *think* some changes to `{parsnip}` are required before this PR...
The `predict()` function has an argument `time` which captures the times at which to predict the survival probability. The functions to do the calculations also have arguments or objects `time`...
Very messy start to the functionality, I couldn't remove this gitnore sorry, also we will eventually have to implement survival:cox but that returns hazard ratios instead of time so I...
the vignette for survival analysis and resampling needs a new home, possibly in censored or tidymodels.org https://github.com/tidymodels/rsample/blob/bafa7ce3f7498896df3b6bfa32632e0858ca9011/vignettes/Applications/Survival_Analysis.Rmd
Apologies if I'm doing something silly here -- or if this is on the roadmap already, or I'm in the wrong place! I can't seem to figure out how to...
make use of rlang 1.0.0 errors and pass the call down to the check functions
With mgcv, example [here](https://rdrr.io/cran/mgcv/man/coxph.html)
With `glmnet` (and maybe `mboost`) we can get predictions one many values of some tuning parameters without multiple model fits. `parsnip` saves those as nested tibbles per row of `new_data`....