getBF {dynaTree}

Extract a Path of (log) Bayes Factors

Description

Extract a path (log) Bayes factors (BFs) from the log marginal posterior probabilities of two "dynaTree"-class objects

Usage

getBF(obj1, obj2)

Arguments

obj1

a "dynaTree"-class object built by dynaTree

obj2

another "dynaTree"-class object built by dynaTree

Details

Simply calculates a difference in log marginal posterior probabilities, setting BFs to zero for initial elements of the path where one of the objects has more zero marginal probabilities than the other. The BF is for the model in obj1 over obj2. If the objects are the output of repeated fits as obtained from dynaTrees, then multiple traces are returned

Value

Returns a vector or matrix of a trace(s) of Bayes factors that can be plotted; see examples below

Author(s)

Robert B. Gramacy [e-mail],
Matt Taddy and Christoforos Anagnostopoulos

References

Taddy, M.A., Gramacy, R.B., and Polson, N. (2011). “Dynamic trees for learning and design” Journal of the American Statistical Association, 106(493), pp. 109-123; arXiv:0912.1586

Gramacy, R.B., Taddy, M.A., and S. Wild (2011). “Variable Selection and Sensitivity Analysis via Dynamic Trees with an Application to Computer Code Performance Tuning” arXiv:1108.4739

https://bobby.gramacy.com/r_packages/dynaTree/

See Also

dynaTree, update.dynaTree, link{logpost}

Examples

## parabola data
n <- 100
Xp <- sort(runif(n,-3,3))
Yp <- Xp + Xp^2 + rnorm(n, 0, .2)
XXp <- seq(-3,3,length=100)

## comparison by log Bayes Factor
R <- 2 ## use R >= 10 for better results
## small R is for faster CRAN checks
o <- apply(matrix(runif(n*(R-1)), ncol=R-1), 2, order)
lpc.p <- dynaTrees(Xp, Yp, R=R, rorder=o, verb=0)
lpl.p <- dynaTrees(Xp, Yp, model="linear", R=R, rorder=o, verb=0)
bf.p <- getBF(lpl.p, lpc.p)

## plot the log Bayes factors
matplot(bf.p, type="l", lty=1, col="gray", main="parabola",
        xlab="time", ylab="log Bayes factor")

## see demo("reg1d") for further examples