Changes in version 10.0 o Switch to fru as a default importance source provider; this will be faster, yet will provide slightly different results due to a stochastic nature of Random Forest. Fru also doesn't support Gini importance, censored data and requires Rust for compilation. To go back to ranger, load it and use getImp=getImpRfZ in the Boruta call. o Direct Xgboost support was dropped; this is not a good idea but one can use custom adapter to restore this functionality. o Ranger extra trees importance adapters were removed, one can restore this by using one of getImpRf* adapters and providing spliture="extratrees" in the Boruta call. o Documentation updates. Changes in version 9.0 o Documentation updates. Changes in version 8.0 o Faster parser for the formula interface. o Transdapter for handling missing values with importance sources which do not support them. o Transdapter for applying decoherence transformation to input. o Transdapter for conditining importance. o Vignette about transdapters. o TentativeRoughFix fixed to work properly with a single Tentative feature. Changes in version 7.0 o Migration from NotABug to GitLab. o Improved testing. o Improved Xgboost adapter. o Xgboost adapter fixed to pass `...` to xgboost function. (Thanks to `ObsoleteVersion` for spotting this) Changes in version 6.0 o Migration from GitHub to NotABug. o Additional verbosity level, doTrace=3. o `Boruta for those in a hurry` vignette. o Extra-trees importance. Changes in version 5.3 o Xgboost importance. o Codebase modernisation. Changes in version 5.2 o New status messages, including proper grammar and a count of still undecided attributes. (Thanks to Aristide Mooyaart for suggesting this) o Boruta to formula conversion functions (getConfirmedFormula, getNonRejectedFormula) are back, after a silent disappearance in the 2.0 version. (Thanks to Marco Niemann) o Documentation fixes. Changes in version 5.1 o Documentation fix. (Thanks to Andrew Slack-Smith for spotting the problem) o ranger VIM providers fixed to support survival problems. (Thanks to James Iremonger for spotting this) o Various documentation updates. o Code and issues on GitHub. Changes in version 5.0 o ranger RF implementation replaced randomForest in default importance providers, bringing speed and parallel processing capabilities. Results should be the same, but there may be minor differences. Using getImp=getImpLegacyRfZ restores legacy behaviour. See ?getImpLegacyRf for a further discussion. o Names of columns in the attStats output changed to reflect the fact that Boruta may use arbitrary importance source. o Documentation fixes, better organisation of code over files. Changes in version 4.0 o Initial rounds were dropped from the algorithm, as they seem redundant after implementation of a proper p-value adjustment. This should provide noticeable speed-up, however the results may be different than in previous versions. o TentativeRoughFix interface changes in order to adapt to the change mentioned above. o Output of the importance source is validated, so that wrong values will be clearly reported. o Code clean-up. Changes in version 3.1 o Better status messages. Changes in version 3.0 o More importance adapters defined. o Fixed TentativeRoughFix() bug making this function always return the original Boruta object without any modification. o Built-in multiple-comparisons correction, which breaks compatibility with older version of Boruta. Use pValue=0.001 and mcAdj=FALSE to trigger the legacy behaviour. o Experimental, "forced" mode of Boruta dropped. It is equivalent to fixing light argument of Boruta function to TRUE, its previously default value. Changes in version 2.1 o Fixed attStats() bug making withTentative argument behave opposite to what it should do. (Thanks to Jitao David Zhang for spotting this) Changes in version 2.0 o Allowed to replace randomForest with other importance source. o Changes in output; "rand" attributes are now called "shadow" ones. o ZScoreHistory is now ImpHistory; some scripts and old Boruta objects may become incompatible. o Documentation updates. o Formulae export functions become private. Changes in version 1.5 (2010-12-15) o Improved performance of dealing with huge inputs. o Removed redundant decisionHistory from output. Changes in version 1.4 (2010-11-26) o Fixed manual entry suggesting that Boruta can only work for classification problems. (Thanks to Allan Engelhardt for spotting this) o Fixed an error occurring when all attributes are claimed Rejected during one of the initial rounds. (Thanks to Allan Engelhardt for spotting this) Changes in version 1.3 (2010-09-02) o Fixes to the output. (Thanks to Achim Zeileis for suggesting them) Changes in version 1.2 (2010-05-08) o New plotting function plotZHistory. o Changes in manual. Changes in version 1.1 (2010-04-18) o Fixed a bug making get*Formula functions throw error when Boruta was run with formula provided as a variable. o Fixed minor documentation inconsistency. Changes in version 1.0 (2009-12-06) o First public release.