Error: message "with parameters optimization using mlr3"

Question

I am using parameters optimization (random search) with mlr3 but it gives me the following error. I tried with other models too (kknn) but the same error comes in.

Error: Resampling 'cv' may not be instantiated

///My code is here 

data = readARFF("xerces.arff")
index= sample(1:nrow(data), 0.7*nrow(data))
train= data[index,]
test= data[-index,]
task = TaskRegr$new("data", backend = train, target = "bug")

learner5=lrn("regr.randomForest")
resampling_cv = rsmp("cv", folds = 10L)
resampling_cv$instantiate(task)
measure= msr("regr.mae")

search_space = paradox::ParamSet$new(
 params = list(paradox::ParamInt$new("ntree", lower = 100, upper = 500)))

terminator = trm("evals", n_evals = 30)
tuner = tnr("random_search")

at = AutoTuner$new(
        learner = learner5,
        resampling = resampling_cv,
        measure = measure,
        search_space = search_space,
        terminator = terminator,
        tuner = tuner,  store_tuning_instance = TRUE,
        store_benchmark_result = TRUE,
        store_models = TRUE
)

be-marc · Accepted Answer · 2022-05-28 17:47:14Z

1

You instantiate the resampling with resampling_cv$instantiate(task). Remove this line and it should work.

answered May 28, 2022 at 17:47

be-marc

1,6217 silver badges6 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Neha Over a year ago

reampling should not be instantiated? mlr3 book says it should be.

be-marc Over a year ago

Usually, the resampling is automatically instantiated. Where did you read that in the book?

Neha Over a year ago

I read it it in the "resamplings" (3rd chapter) Performance Evaluation and Comparison

be-marc Over a year ago

Yes it says: Note that if you want to compare multiple Learners in a fair manner, using the same instantiated resampling for each learner is mandatory, such that each learner gets exactly the same training data and the performance of the trained model is evaluated in exactly the same test set.

be-marc Over a year ago

If you want to compare multiple learners, use nested resampling. You don't have to instantiate the resampling before.

Collectives™ on Stack Overflow

Error: message "with parameters optimization using mlr3"

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related