This C++ version of BAT is still being maintained, but addition of new features is unlikely. Check out our new incarnation, BAT.jl, the Bayesian analysis toolkit in Julia. In addition to Metropolis-Hastings sampling, BAT.jl supports Hamiltonian Monte Carlo (HMC) with automatic differentiation, automatic prior-based parameter space transformations, and much more. See the BAT.jl documentation.

Known issues in version 0.2

• When reading data from a text file using the  BCDataSet::ReadDataFromFileTxt method no empty line is allowed at the end of the file. Otherwise the last point is included twice in the data set.
• The MCMC sampling near boundary of the parameter space can be biased if the bulk of the probability distribution lies very close to this boundary, e.g. for parameter between 0 and 100 with p=1 between 0 and 1 and p=0 between 1 and 100 gives non-flat sampling below 1. This doesn't arise if range is small, e.g. for the case above range 0-2 would be OK.