NEWS

---

rFerns 5.0

• Fixed clang-13 issue with OpenMP.
• Changed HTTP URLs to use HTTPS.

rFerns 4.0

• Internal PRNG switched to PCG32, with D. Lemire method for used for unbiased sampling. Hence, same random seed does not reproduce the same results as pre-4.0 versions. One can also expect slight speed regression.
• Migrated to GitLab.
• Fixed error in NA checking. Thanks to Patrick Schratz for spotting this.

rFerns 3.0

• Parallel training with OpenMP.
• New organisation of the internal PRNG.
• Error propagation reporting removed (arguments reportErrorEvery and saveErrorPropagation) since it brings too little value for the complexity it requires. And can be emulated with merging anyway.

rFerns 2.0.3 (2017-11-15)

• Registration of native routines.
• Documentation updates.

rFerns 2.0.2 (2016-07-22)

• Added a wrapper for performing feature selection using shadow importance.

rFerns 2.0.1 (2016-05-20)

• Fixed bug in merge.rFerns code, causing merged importance scores to be wrong, especially after merging very small ensembles.
• Fixed memory protection bug triggering rare segmentation faults in the R's GC, often long after the rFerns model was completed.

rFerns 2.0.0 (2015-12-03)

• Shadow importance.
• Roxygen documentation.
• Model merging for parallel, batch and deeply traceable training.
• Fixed a minor error in leaf score calculation (basically redundant to the model performance). Thanks to Mateusz Fedoryszak for spotting this.
• Fixed OOB predictions returned by predict(), which are now always data.frame, even in the multi-label case.

rFerns 1.1.0 (2014-11-30)

• Code portability fixes.
• predict method returns OOB results for no input data even when fern forest is not saved.
• Fixed minor importance calculation bug.

rFerns 1.0.0 (2014-11-11)

• Published in JSS.

rFerns 0.3.3 (2014-03-24)

• Fixed uninitialised memory bug causing problems with LLVM.

rFerns 0.3.2 (2014-02-18)

• Prototype of a multi-label classification.
• Removed redundant constant from scores definition; score values will be different, but the importance scores and prediction results will be the same as generated by previous versions.

rFerns 0.3.1 (2013-04-23)

• predict method for rFerns returns OOB votes/scores when no predictors are given.

rFerns 0.3 (2012-01-18)

• Handling of unbalanced classes.
• Performance tweaks.

rFerns 0.2 (2012-01-02)

• Added support for integer and ordered-factor attributes, suggested by Bill Venables.

rFerns 0.1 (2011-12-30)

• First public release.