Classification Learner for Debugging

A simple LearnerClassif used primarily in the unit tests and for debugging purposes. If no hyperparameter is set, it simply constantly predicts a randomly selected label. The following hyperparameters trigger the following actions:

error_predict:

Probability to raise an exception during predict.

error_train:

Probability to raises an exception during train.

message_predict:

Probability to output a message during predict.

message_train:

Probability to output a message during train.

predict_missing:

Ratio of predictions which will be NA.

predict_missing_type:

To to encode missingness. “na” will insert NA values, “omit” will just return fewer predictions than requested.

save_tasks:

Saves input task in model slot during training and prediction.

segfault_predict:

Probability to provokes a segfault during predict.

segfault_train:

Probability to provokes a segfault during train.

sleep_train:

Function returning a single number determining how many seconds to sleep during $train().

sleep_predict:

Function returning a single number determining how many seconds to sleep during $predict().

threads:

Number of threads to use. Has no effect.

warning_predict:

Probability to signal a warning during predict.

warning_train:

Probability to signal a warning during train.

x:

Numeric tuning parameter. Has no effect.

iter:

Integer parameter for testing hotstarting.

count_marshaling:

If TRUE, marshal_model will increase the marshal_count by 1 each time it is called. The default is FALSE.

Note that segfaults may not be triggered reliably on your operating system. Also note that if they work as intended, they will tear down your R session immediately!

Dictionary

This Learner can be instantiated via the dictionary mlr_learners or with the associated sugar function lrn():

mlr_learners$get("classif.debug")
lrn("classif.debug")

Meta Information

• Task type: “classif”

• Predict Types: “response”, “prob”

• Feature Types: “logical”, “integer”, “numeric”, “character”, “factor”, “ordered”

• Required Packages: mlr3

Parameters

Id	Type	Default	Levels	Range
error_predict	numeric	0		\([0, 1]\)
error_train	numeric	0		\([0, 1]\)
message_predict	numeric	0		\([0, 1]\)
message_train	numeric	0		\([0, 1]\)
predict_missing	numeric	0		\([0, 1]\)
predict_missing_type	character	na	na, omit	-
save_tasks	logical	FALSE	TRUE, FALSE	-
segfault_predict	numeric	0		\([0, 1]\)
segfault_train	numeric	0		\([0, 1]\)
sleep_train	untyped	-		-
sleep_predict	untyped	-		-
threads	integer	-		\([1, \infty)\)
warning_predict	numeric	0		\([0, 1]\)
warning_train	numeric	0		\([0, 1]\)
x	numeric	-		\([0, 1]\)
iter	integer	1		\([1, \infty)\)
count_marshaling	logical	FALSE	TRUE, FALSE	-

See also

• Chapter in the mlr3book: https://mlr3book.mlr-org.com/chapters/chapter2/data_and_basic_modeling.html#sec-learners

• Package mlr3learners for a solid collection of essential learners.

• Package mlr3extralearners for more learners.

• Dictionary of Learners: mlr_learners

• as.data.table(mlr_learners) for a table of available Learners in the running session (depending on the loaded packages).

• mlr3pipelines to combine learners with pre- and postprocessing steps.

• Package mlr3viz for some generic visualizations.

• Extension packages for additional task types:

  • mlr3proba for probabilistic supervised regression and survival analysis.

  • mlr3cluster for unsupervised clustering.

• mlr3tuning for tuning of hyperparameters, mlr3tuningspaces for established default tuning spaces.

Other Learner: Learner, LearnerClassif, LearnerRegr, mlr_learners, mlr_learners_classif.featureless, mlr_learners_classif.rpart, mlr_learners_regr.debug, mlr_learners_regr.featureless, mlr_learners_regr.rpart

Super classes

mlr3::Learner -> mlr3::LearnerClassif -> LearnerClassifDebug

Active bindings

marshaled

(logical(1))
Whether the learner has been marshaled.

Methods

Public methods

• LearnerClassifDebug$new()

• LearnerClassifDebug$marshal()

• LearnerClassifDebug$unmarshal()

• LearnerClassifDebug$clone()

Inherited methods

• mlr3::Learner$base_learner()
• mlr3::Learner$format()
• mlr3::Learner$help()
• mlr3::Learner$predict()
• mlr3::Learner$predict_newdata()
• mlr3::Learner$print()
• mlr3::Learner$reset()
• mlr3::Learner$train()

---

Method new()

Creates a new instance of this R6 class.

Usage

LearnerClassifDebug$new()

---

Method marshal()

Marshal the learner's model.

Usage

LearnerClassifDebug$marshal(...)

Arguments

...

(any)
Additional arguments passed to marshal_model().

---

Method unmarshal()

Unmarshal the learner's model.

Usage

LearnerClassifDebug$unmarshal(...)

Arguments

...

(any)
Additional arguments passed to unmarshal_model().

---

Method clone()

The objects of this class are cloneable with this method.

Usage

LearnerClassifDebug$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

learner = lrn("classif.debug")
learner$param_set$values = list(message_train = 1, save_tasks = TRUE)

# this should signal a message
task = tsk("penguins")
learner$train(task)
#> Message from classif.debug->train()
learner$predict(task)
#> <PredictionClassif> for 344 observations:
#>     row_ids     truth response
#>           1    Adelie   Gentoo
#>           2    Adelie   Gentoo
#>           3    Adelie   Gentoo
#> ---                           
#>         342 Chinstrap   Gentoo
#>         343 Chinstrap   Gentoo
#>         344 Chinstrap   Gentoo

# task_train and task_predict are the input tasks for train() and predict()
names(learner$model)
#> [1] "response"      "pid"           "iter"          "id"           
#> [5] "random_number" "task_train"    "task_predict"