Bad binding access while using MLR within a loop

[gabrielsaul]

Getting an error while parallel training an ML model inside a loop. The errors vary and they seem to happen at random during the parallel training phase. Here is one such example:

The full code can be found here

This is the code wherein the error occurs:

      credit.task = makeClassifTask(id = "credit",
                                    data = credit, target = "risk")
      lrn = makeLearner("classif.svm", predict.type = "prob")
      credit.lrn = cpoScale() %>>% cpoDummyEncode() %>>% lrn
      param.set = pSS(
        cost: numeric[0.01, 1]
      )
      
      TUNEITERS = 100L
      RESAMPLING = cv5
      
      if (PARALLEL) {
        parallelMap::parallelStartSocket(parallel::detectCores(), level = "mlr.tuneParams")
      }
      
      ctrl = makeTuneControlRandom(maxit = TUNEITERS * length(param.set$pars))
      lrn.tuning = makeTuneWrapper(lrn, RESAMPLING, list(mlr::acc), param.set, ctrl, show.info = FALSE)
      res = tuneParams(lrn, credit.task, RESAMPLING, par.set = param.set, control = ctrl,
                       show.info = FALSE)
      performance = resample(lrn.tuning, credit.task, RESAMPLING, list(mlr::acc))$aggr
      
      if (PARALLEL) {
        parallelMap::parallelStop()
      }
      credit.lrn = setHyperPars2(credit.lrn, res$x) 
      credit.model = mlr::train(credit.lrn, credit.task)

NOTE: The exact same code works perfectly fine when executed outside of a loop.

NOTE: This issue started appearing after I updated my RStudio to the current latest version (Version: 2023.12.0+369)

NOTE: Strangely, if I place all of the code inside of a function and then call said function, it works again.

This makes me think it's a nesting issue, but I have no idea. I am relatively new to R and I am merely reviewing existing code for use in my research.

Any guidance will be greatly appreciated. Thank you.

[mikecrobp]

I had the same issue after upgrading to 2023.12.0+369.
I was using httr
Reverting to previous version of Rstudio fixed it
I don't have a reproducible example yet

[gabrielsaul]

Have you tried placing it inside a function and executing it that way?