optimizing of tuning instance does not work

[ElsaVazquez29]

I had to install the newest version of mlr3 and its dependencies because my computer was updated.

The code below used to work just fine.

#tune random forest model
learner_rfsrc=lrn('surv.rfsrc', mtry=to_tune(7,50), nodesize=to_tune(5, 50), ntree=to_tune(500,1000), importance='permute')
learner_rfsrc$param_set
#declare tuning instance for random survival forest
ti_rfsrc=ti(task=tsk_chf_training_new,
learner=learner_rfsrc,
resampling=rsmp("cv", folds=5),
measures=msr("surv.cindex"),
terminator=trm("evals", n_evals=50))

#tune random forest
tuner_random$optimize(ti_rfsrc)

Now we I try to optimize the tuning instance I get the following error

INFO [10:58:06.731] [bbotk] Starting to optimize 3 parameter(s) with '' and ' [n_evals=50, k=0]'
INFO [10:58:06.940] [bbotk] Evaluating 1 configuration(s)
Error in if (resampling$task_row_hash != task$row_hash) { :
argument is of length zero

[larskotthoff]

Can you provide a complete example that allows to reproduce the problem please?

[be-marc]

If the issue persist after updating your packages, please reopen.