RaphaelS1 / mlr3extralearners

Extra learners for use in mlr3.

mlr3extralearners

Package website: release | dev

Extra Learners for mlr3.

What is mlr3extralearners?

mlr3extralearners contains all learners from mlr3 that are not in mlr3learners or the core packages. mlr3extralearners contains helper functions to find where all the learners, across the mlr3verse, live and to install required packages to run these learners. See the interactive learner list for the full list of learners in the mlr3verse and the learner status page for a live build status.

list_mlr3learners(select = c("id", "mlr3_package", "required_packages"))
#>                      id      mlr3_package      required_packages
#>   1: classif.AdaBoostM1 mlr3extralearners                  RWeka
#>   2:        classif.C50 mlr3extralearners                    C50
#>   3:    classif.cforest mlr3extralearners partykit,sandwich,coin
#>   4:      classif.ctree mlr3extralearners partykit,sandwich,coin
#>   5:  classif.cv_glmnet      mlr3learners                 glmnet
#>  ---                                                            
#>  97:      surv.pchazard mlr3extralearners      reticulate,pracma
#>  98:     surv.penalized mlr3extralearners              penalized
#>  99:         surv.rfsrc mlr3extralearners        randomForestSRC
#> 100:         surv.rpart         mlr3proba  rpart,distr6,survival
#> 101:           surv.svm mlr3extralearners            survivalsvm

mlr3extralearners lives on GitHub and will not be on CRAN. Install with:

remotes::install_github("mlr-org/mlr3extralearners")

Installing and Loading Learners

The package includes functionality for detecting if you have the required packages installed to use a learner, and ships with the function install_learner which can install all required learner dependencies.

lrn("regr.gbm")
#> Error: Required packages not installed, please run `install_learners("regr.gbm")`.

install_learners("regr.gbm")

lrn("regr.gbm")
#> <LearnerRegrGBM:regr.gbm>
#> * Model: -
#> * Parameters: keep.data=FALSE, n.cores=1
#> * Packages: gbm
#> * Predict Type: response
#> * Feature types: integer, numeric, factor, ordered
#> * Properties: importance, missings, weights

Extending mlr3extralearners

New learners can be created with the create_learner function. This assumes you have a local copy of mlr3extralearners. This function will automatically create the learner, learner tests, parameter tests, YAML files for CI if required, and update the DESCRIPTION if required. Once all tests are passing locally, open a pull request with the “New Learner” template.

create_learner(classname = "Locfit",
               algorithm = "localised fit",
               type = "dens",
               key = "locfit",
               package = "locfit",
               caller = "density.lf",
               feature_types = c("integer", "numeric"),
               predict_types = c("pdf", "cdf"),
               properties = NULL,
               importance = FALSE,
               oob_error = FALSE,
               references = FALSE)