Bactrian random walk operator
A random walk operator that selects a random dimension of the real parameter and perturbs the value a random amount according to a Bactrian distribution (Yang & Rodriguez, 2013), which is a mixture of two Gaussians:p(x) = 1/2N(x;-m,1-m^2) + 1/2N(x;+m,1-m^2) and more efficient than non-Bactrian RealRandomWalkOperator
The parameter to operate a random walk on. (required)
Scaling factor: larger means more bold proposals.
optimise is true, this is automatically tuned during MCMC (recommended).
Flag to indicate that the scale factor is automatically changed in order to achieve a good acceptance rate (default true)
Provides sample distribution for proposals (optional, default: bactrian) See also kernel distributions.
Weight with which this operator is selected (required). See also operator weight tuning.
Yang Z, Rodríguez CE. Searching for efficient Markov chain Monte Carlo proposal kernels. Proceedings of the National Academy of Sciences. 2013 Nov 26;110(48):19307-12. doi: 10.1073/pnas.1311790110.