Michel Lang

Functions introduced or changed since R v3.0.0 are re-implemented in this package. The backports are conditionally exported in order to let R resolve the function name to either the implemented backport, or the respective base version, if available. Package developers can make use of new functions or arguments by selectively importing specific backports to support older installations.

In contrast to RFC3548, the 62nd character ("+") is replaced with "-", the 63rd character ("/") is replaced with "_". Furthermore, the encoder does not fill the string with trailing "=". The resulting encoded strings comply to the regular expression pattern "[A-Za-z0-9_-]" and thus are safe to use in URLs or for file names. The package also comes with a simple base32 encoder/decoder suited for case insensitive file systems.

Extends the BatchJobs package to run statistical experiments on batch computing clusters. For further details see the project web page.

As a successor of the packages 'BatchJobs' and 'BatchExperiments', this package provides a parallel implementation of the Map function for high performance computing systems managed by schedulers 'IBM Spectrum LSF' (<https://www.ibm.com/us-en/marketplace/hpc-workload-management>), 'OpenLava' (<http://www.openlava.org/>), 'Univa Grid Engine'/'Oracle Grid Engine' (<http://www.univa.com/>), 'Slurm' (<http://slurm.schedmd.com/>), 'TORQUE/PBS' (<https://adaptivecomputing.com/cherry-services/torque-resource-manager/>), or 'Docker Swarm' (<https://docs.docker.com/swarm/>). A multicore and socket mode allow the parallelization on a local machines, and multiple machines can be hooked up via SSH to create a makeshift cluster. Moreover, the package provides an abstraction mechanism to define large-scale computer experiments in a well-organized and reproducible way.

Tests and assertions to perform frequent argument checks. A substantial part of the package was written in C to minimize any worries about execution time overhead.

Implements specialized conditions, i.e., typed errors, warnings and messages. Offers a set of standardized conditions (value error, deprecated warning, io message, ...) in the fashion of Python's built-in exceptions.

More comfortable interface to work with R data or source files in a key-value fashion.

Efficient, object-oriented programming on the building blocks of machine learning. Provides 'R6' objects for tasks, learners, resamplings, and measures. The package is geared towards scalability and larger datasets by supporting parallelization and out-of-memory data-backends like databases. While 'mlr3' focuses on the core computational operations, add-on packages provide additional functionality.

A small collection of interesting and educational machine learning data sets which are used as examples in the 'mlr3' book (<https://mlr3book.mlr-org.com>), the use case gallery (<https://mlr3gallery.mlr-org.com>), or in other examples. All data sets are properly preprocessed and ready to be analyzed by most machine learning algorithms. Currently contains the following data sets: (1) housing prices in Kings County, and (2) Titanic passenger survival data.

Extends the 'mlr3' package with a backend to transparently work with data bases. Internally relies on the abstraction of package 'dbplyr' to interact with one of the many supported data base management systems (DBMS).

Recommended Learners for 'mlr3'. Extends 'mlr3' with interfaces to essential machine learning packages on CRAN. This includes, but is not limited to: (penalized) linear and logistic regression, linear and quadratic discriminant analysis, k-nearest neighbors, naive Bayes, support vector machines, and gradient boosting.

Frequently used helper functions and assertions used in 'mlr3' and its companion packages. Comes with helper functions for functional programming, for printing, to work with 'data.table', as well as some generally useful 'R6' classes. This package also supersedes the package 'BBmisc'.

Implements methods for hyperparameter tuning with 'mlr3', e.g. Grid Search, Random Search, or Simulated Annealing. Various termination criteria can be set and combined. The class 'AutoTuner' provides a convenient way to perform nested resampling in combination with 'mlr3'.

Provides visualizations for 'mlr3' objects such as tasks, predictions, resample results or benchmark results via the autoplot() generic of 'ggplot2'. The returned 'ggplot' objects are intended to provide sensible defaults, yet can easily be customized to create camera-ready figures. Visualizations include barplots, boxplots, histograms, ROC curves, and Precision-Recall curves.

Define parameter spaces, constraints and dependencies for arbitrary algorithms, to program on such spaces. Also includes statistical designs and random samplers. Objects are implemented as 'R6' classes.

Provides Map, Reduce and Filter variants to generate jobs on batch computing systems like PBS/Torque, LSF, SLURM and Sun Grid Engine. Multicore and SSH systems are also supported. For further details see the project web page.

Miscellaneous helper functions for and from B. Bischl and some other guys at TU Dortmund, mainly for package development.

Fast aggregation of large data (e.g. 100GB in RAM), fast ordered joins, fast add/modify/delete of columns by group using no copies at all, list columns, friendly and fast character-separated-value read/write. Offers a natural and flexible syntax, for faster development.

Parsing and evaluation tools that make it easy to recreate the command line behaviour of R.

This package provides modified versions and novel implementation of functions for parallel evaluation, tailored to use with Bioconductor objects.

Interface to a large number of classification and regression techniques, including machine-readable parameter descriptions. There is also an experimental extension for survival analysis, clustering and general, example-specific cost-sensitive learning. Generic resampling, including cross-validation, bootstrapping and subsampling. Hyperparameter tuning with modern optimization techniques, for single- and multi-objective problems. Filter and wrapper methods for feature selection. Extension of basic learners with additional operations common in machine learning, also allowing for easy nested resampling. Most operations can be parallelized.

Extends 'mlr3' with filter methods for feature selection. Besides standalone filter methods built-in methods of any machine-learning algorithm are supported. Partial scoring of multivariate filter methods is supported.

Dataflow programming toolkit that enriches 'mlr3' with a diverse set of pipelining operators ('PipeOps') that can be composed into graphs. Operations exist for data preprocessing, model fitting, and ensemble learning. Graphs can themselves be treated as 'mlr3' 'Learners' and can therefore be resampled, benchmarked, and tuned.

Provides extensions for probabilistic supervised learning for 'mlr3'. This includes extending the regression task to probabilistic and interval regression, adding a survival task, and other specialized models, predictions, and measures.

The 'mlr3' package family is a set of packages for machine-learning purposes built in a modular fashion. This wrapper package is aimed to simplify the installation and loading of the core 'mlr3' packages. Get more information about the 'mlr3' project at <https://mlr3book.mlr-org.com/>.

Toolset that enriches 'mlr' with a diverse set of preprocessing operators. Composable Preprocessing Operators ("CPO"s) are first-class R objects that can be applied to data.frames and 'mlr' "Task"s to modify data, can be attached to 'mlr' "Learner"s to add preprocessing to machine learning algorithms, and can be composed to form preprocessing pipelines.

Flexible and comprehensive R toolbox for model-based optimization ('MBO'), also known as Bayesian optimization. It implements the Efficient Global Optimization Algorithm and is designed for both single- and multi- objective optimization with mixed continuous, categorical and conditional parameters. The machine learning toolbox 'mlr' provide dozens of regression learners to model the performance of the target algorithm with respect to the parameter settings. It provides many different infill criteria to guide the search process. Additional features include multi-point batch proposal, parallel execution as well as visualization and sophisticated logging mechanisms, which is especially useful for teaching and understanding of algorithm behavior. 'mlrMBO' is implemented in a modular fashion, such that single components can be easily replaced or adapted by the user for specific use cases.

We provide an R interface to 'OpenML.org' which is an online machine learning platform where researchers can access open data, download and upload data sets, share their machine learning tasks and experiments and organize them online to work and collaborate with other researchers. The R interface allows to query for data sets with specific properties, and allows the downloading and uploading of data sets, tasks, flows and runs. See <https://www.openml.org/guide/api> for more information.

Unified parallelization framework for multiple back-end, designed for internal package and interactive usage. The main operation is parallel mapping over lists. Supports 'local', 'multicore', 'mpi' and 'BatchJobs' mode. Allows tagging of the parallel operation with a level name that can be later selected by the user to switch on parallel execution for exactly this operation.

Functions for parameter descriptions and operations in black-box optimization, tuning and machine learning. Parameters can be described (type, constraints, defaults, etc.), combined to parameter sets and can in general be programmed on. A useful OptPath object (archive) to log function evaluations is also provided.

Robust and efficient feature selection algorithm to identify important features for predicting survival risk. The method is based on subsampling and averaging linear models obtained from the (preconditioned) Lasso algorithm, with an extra shrinking procedure to reduce the size of signatures. An evaluation procedure using subsampling is also provided.