Version 5.0.2 of BayesTraits Bug Fix - The chains state would be set to stepping stones, after burn in, if an infinite number of iterations set. Incorrect information about the format of binary maps has been removed. Version 5.0.1 of BayesTraits Bug Fix - GLM models have tree transforms applied twice, since V4.0.0. GLM and contrast results are now the same Version 5.0.0 of BayesTraits Added - Farbic regression model. More to follow Version 4.1.3 of BayesTraits Bug Fix - fixed a bug when interpolating data with the discrete (independent and dependent) model. When redirecting the standard input the file line ending can be Windows / Linux / OS X. Version 4.1.2 of BayesTraits Bug Fix - fixed a crash when interpolating data with the GLM regression model. Removed the ability to interpolate independent data or estimate ancestral states with the GLM regression model. Version 4.1.1 of BayesTraits Bug Fix - proposal mechanism issue with the fabric betas x branch (VR_LS_BL) parameters. Removed the default prior on fabric betas x branch lengths (VR_LS_BL) as the prior is data specific. Ancestral states can now be reconstructed for fabric and global trend models. The default prior on the node scalars (fabric / variable rates) has been changes to its earlier distribution. Version 4.0.1 of BayesTraits Bug Fix - Geo Model - proposal mechanism was biased around the Antimeridian. Version 4.0.0 of BayesTraits Automatically detects and models directional changes, as well as evolvability changes. See Pagel etal, General statistical model shows that macroevolutionary patterns and processes are consistent with Darwinian gradualism, Nature Communications, 2022 A global trend model to test if trait change shows a general pattern of increase or decrease along paths through the tree that differ in their total evolvability. Variable Rates models and other Reverse Jump tree transformation methods can be used on samples of trees, incorporating uncertainty in topology and branch length estimates into the analysis. Threshold costs can be applied to Reverse Jump tree transformations Draw random values from prior distributions, which can be used to visualise priors Speed improvements when using distributions of data, for large trees this can improve performance by two orders of magnitude. Threaded version of the Geo model and non-threaded are consistent with each other for a given random seed. Version 3.0.5 of BayesTraits Bug Fix - Geo Model with Distribution of data - updating ancestral sates could use old tip values. Bug Fix - Geo Model - an incorrect likelihood could be accepted when the RNG produces a zero value. Change - Zero RJ rates was relaxed from 1E-21 to 1E-07. Add - Prior for CV switching rates. Version 3.0.2 of BayesTraits Bug Fix - Maximum likelihood values of kappa and gamma were incorrect. Change - Geo model can now use variable rates. Version 3.0.1 of BayesTraits Bug Fix - Fossilising state 1,0 was set to the fossil value of state 0,1 with discrete models. Version 3.0.0 of BayesTraits Automatically detect shifts in rates of evolution (Variable Rates model) for MultiState / discrete data, as well as continuous. Kappa, lambda, delta and rate scalars can be applied to nodes within a tree Allow patterns of evolution to vary within a tree for MultiState / discrete data Improved parallelism Integration of a fast / high precision likelihood calculation for multi-state and discrete models Reverse Jump MCMC (RJ-MCMC) methods to detect changes in evolutionary patterns (kappa, lambda, and delta) Improved Maximum Likelihood searching Geographical models Distributions of trait data instead of single values More control over priors