How to define gamma.mu for surv.svm in paradox?

[wangbaili]

I'm new to mlr3 and R.I'm try to adjust parameter for surv.svm, but I can define the ParamSpace by paradox,because the type of gamma.mu is ParamFct. This is my code:

svm<- lrn("surv.svm")

ps = ParamSet$new(params = list(ParamDbl$new(id = 'bound',lower = 5, upper=15), ParamFct$new(id = 'opt.meth',levels = c("quadprog","ipop")), ParamDbl$new(id = 'margin',lower = 0, upper=0.2), ParamUty$new(id='gamma.mu'),(how to code gamma.mu?) ParamFct$new(id = 'kernel',levels = c("lin_kernel","add_kernel","rbf_kernel","poly_kernel")))) at= AutoTuner$new(learner =svm, resampling = rsmp("holdout"), measure = msr("surv.cindex"), terminator = trm("evals", n_evals = 5), tuner = tnr("random_search"), search_space=ps)

Thank you！

[bblodfon]

Hi @wangbaili,

The code in this issue answers your question. Note that as I report there, I would advise against using survival SVMs.