Skip to contents

Changelog

Source: NEWS.md

mlr3benchmark 0.1.5

CRAN release: 2023-03-14

  • Rename argument of autoplot function to comply with new CRAN check
  • Fix bug in rank_data caused by learner name mismatch

mlr3benchmark 0.1.4

CRAN release: 2022-10-10

  • Add friedman_global argument to posthoc tests and to autoplots to allow methods and plots to run even if the global Friedman test fails (i.e. don’t reject null)
  • New maintainer: Sebastian Fischer
  • Fix documentation

mlr3benchmark 0.1.3

CRAN release: 2021-10-04

  • Fix README
  • Fix for PMCMRplus

mlr3benchmark 0.1.2

CRAN release: 2021-04-19

  • Critical patch for bug in creating BenchmarkAggr objects.
  • Task and learner columns must now be provided as factors to BenchmarkAggr objects, no internal coercion is made.
  • Bug fix in CD plots

mlr3benchmark 0.1.1

CRAN release: 2020-12-16

  • BenchmarkAggr$friedman_test now returns the full test object if only a single measure exists in the object
  • Fixed plotting in autoplot.BenchmarkAggr for CD-plots, previously bars were overlapping and giving misleading results.
  • BenchmarkAggr is now more flexible in construction. Instead of being forced to name columns task_id and learner_id, instead any name can be used if passed to the respective arguments in the constructor.
  • Adds $subset public method to BenchmarkAggr as a thin wrapper around subset for data.table. Returns subsetted data.table.

mlr3benchmark 0.1.0

CRAN release: 2020-11-19

  • Initial CRAN release