Forum: help
Monitor Forum | 
Start New Thread
| RE: Error in logLik.constparty(...) : not yet implemented [ Reply ] By: Achim Zeileis on 2024-10-25 00:08 | [forum:49996] |
|
Yes. For numeric and categorical data, there are natural log-likelihoods that lead to the same predictions one would usually use. But the same is not true for survival data, unfortunately. |
|
| Error in logLik.constparty(...) : not yet implemented [ Reply ] By: J Kyle Armstrong on 2024-10-24 21:19 | [forum:49995] |
| I noticed that when .response_class(y) is "Surv" that there is no implementation for `varimp` ? | |
