[R] Error in FUN(X[[i]], ...) : subscript out of bounds

[William Dunlap]

In your fit_bayes function, you have
   getTrainPerf(mod)[, "TrainRMSE"]
What are the column name of the output of getTrainPerf(mod)?
   print(colnames(getTrainPerf(mod)))
You can home in on the problem faster if you call
   traceback()
immediately after the error.

Bill Dunlap
TIBCO Software
wdunlap tibco.com


On Wed, Dec 4, 2019 at 3:38 AM Neha gupta <neha.bologna90 using gmail.com> wrote:

> Hi , I know nobody will respond to my query as I asked few in the past and
> no answer received for any of my questions. However, I am asking here with
> the hope it will get responded.
>
> I am using bayesian optimization to tune the parameter of mtry for random
> forest but it gives me the error: Error in FUN(X[[i]], ...) : subscript out
> of bounds
>
> I am using the following code:
>
> fit_bayes <- function(mtry) {
>   ## Use the same model code but for a single (C, sigma) pair.
>   mod <- train(Effort ~ ., data = tr,
>                method = "rf",
>                preProc = c("center", "scale", "zv"),
>                metric = "RMSE",
>                trControl = ctrl,
>                tuneGrid = data.frame(C = 10^(mtry)))
>
>   list(Score = -getTrainPerf(mod)[, "TrainRMSE"], Pred = 0)
> }
>
>
> library(rBayesianOptimization)
>
> bounds <- list(mtry = c(-2,  5)
>
>
> set.seed(8606)
> bo_search <- BayesianOptimization(fit_bayes,
>                                   bounds = bounds,
>                                   init_points = 10,
>                                   n_iter = 100,
>                                   acq = "ucb",
>                                   kappa = 1,
>                                   eps = 0.0)
> bo_search
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

	[[alternative HTML version deleted]]