Add support for approximate Gray log rank split rule #14

Open
opened 2019-04-30 17:39:32 +00:00 by joel · 1 comment
Owner

This is for the case when the data is not censor-complete. Note that it's not clear exactly how randomForestSRC calculates it; but I've verified that they don't use the same maximum observed time in the data prior to making the split so I'll next try recalculating the max for each side of the split.

This is for the case when the data is not censor-complete. Note that it's not clear exactly how `randomForestSRC` calculates it; but I've verified that they don't use the same maximum observed time in the data prior to making the split so I'll next try recalculating the max for each side of the split.
Author
Owner

I've tried several possible combinations of how I think randomForestSRC does it, but none of them match. I've emailed Ishwaran but he didn't immediately know.

This probably isn't going to get implemented anytime soon; if anyone besides me is reading these issues I invite you to implement it and I'd be happy to provide support.

I've tried several possible combinations of how I think randomForestSRC does it, but none of them match. I've emailed Ishwaran but he didn't immediately know. This probably isn't going to get implemented anytime soon; if anyone besides me is reading these issues I invite you to implement it and I'd be happy to provide support.
Sign in to join this conversation.
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: joel/largeRCRF-Java#14
No description provided.