Error in advanced tuning chapter

[larskotthoff]

Running the first chunk of code in chapter 5:

tsk_pen = tsk("penguins")
# remove rows with missing values
tsk_pen$filter(tsk_pen$row_ids[complete.cases(tsk_pen$data())])
# create custom resampling with new factors in test data
rsmp_custom = rsmp("custom")
rsmp_custom$instantiate(tsk_pen,
  list(tsk_pen$row_ids[tsk_pen$data()$island != "Torgersen"]),
  list(tsk_pen$row_ids[tsk_pen$data()$island == "Torgersen"])
)
msr_ce = msr("classif.ce")
tnr_random = tnr("random_search")
learner = lrn("classif.lda", method = "t", nu = to_tune(3, 10))

tune(tnr_random, tsk_pen, learner, rsmp_custom, msr_ce, 10)

I get the error:

Error in lda.default(x, grouping, ...) : 
  variable 6 appears to be constant within groups

[be-marc]

This chunk is supposed to throw an error.
The chunk definition looks like this {r, error = TRUE, warning = FALSE, message = FALSE}.
Do you get an error when running quarto render?
If so, the options may no longer be supported and we have to use #| error: true?

[larskotthoff]

Ah, I see. Yes, I'm getting an error when I render this with quarto.

[be-marc]

Can you try

{r}
#| error: true
#| warning: false
#| message: false

instead of

{r, error = TRUE, warning = FALSE, message = FALSE}

[larskotthoff]

Ok, ran this again -- sorry, the actual error is coming from the chunk after that:

Error in `learner$encapsulate = c(train = "evaluate", predict = "evaluate")`:
! cannot change value of locked binding for 'encapsulate'
Backtrace:
  1. global .main()
  2. execute(...)
  3. rmarkdown::render(...)
  4. knitr::knit(knit_input, knit_output, envir = envir, quiet = quiet)
  5. knitr:::process_file(text, output)
     ...
 16. base::withRestarts(...)
 17. base (local) withRestartList(expr, restarts)
 18. base (local) withOneRestart(withRestartList(expr, restarts[-nr]), restarts[[nr]])
 19. base (local) docall(restart$handler, restartArgs)
 21. evaluate (local) fun(base::quote(`<smplErrr>`))



Quitting from lines 58-59 [optimization-035] (advanced_tuning_methods_and_black_box_optimization.qmd)
Execution halted

So the error is "caught" as expected, but the example afterwards to demonstrate how to do this with encapsulation fails.

[be-marc]

Ah sorry. It's clear that you also have these errors. This is all fixed in #829. @sebffischer Maybe you want to merge?

[sebffischer]

I will meet with bernd on thursday to review the chapter again