Using wattage for survival analysis (random forest forest)

Question

Is there a way to use caret for survival analysis. I really like how easy it is to use. I tried to set a random survival forest using the party package, which is on the carriage list.

It works:

 library(survival) library(caret) library(party) fitcforest <- cforest(Surv(futime, death) ~ sex+age, data=flchain, controls = cforest_classical(ntree = 1000)) 

but using caret , I get an error:

 fitControl <- trainControl(## 10-fold CV method = "repeatedcv", number = 10, repeats = 2, ) cforestfit <- train(Surv(futime, death) ~ sex+age,data=flchain, method="cforest",trControl = fitControl) 

I get this error:

 Error: nrow(x) == length(y) is not TRUE 

Is there any way to get these Surv objects Surv work with the carriage? Can I use other survival-oriented carriage packages?

thanks

Answer 1

Not yet. This is one of two major updates that should be available soon (the other extends preprocessing).

Contact me offline if you are interested in developing and / or testing these features.

Thanks,

Max

Answer 2

I have not found a way to train carriage-based survival models. As an alternative, the mlr framework (1) has a set of students studying survival (2). I found mlr extremely user friendly and useful.

• mlr: http://mlr-org.imtqy.com/mlr-tutorial/release/html/
• Survival students at mlr: http://mlr-org.imtqy.com/mlr-tutorial/release/html/integrated_learners/index.html#survival-analysis-15

Answer 3

In R, the number of packets in which survival data is simulated is growing, examples;

For lasso and stretch nets: BioSpear.

For a random forest: randomForestSRC.

Best, Loick