This package interfaces the BayesX (https://www.uni-goettingen.de/de/bayesx/550513.html)
command-line binary from R. The main model fitting function is called bayesx
.
Before STAR models can be estimated, the command-line version of BayesX needs to be
installed, which is done by installing the R source code package BayesXsrc. Please see
function bayesx
and bayesx.control
for more details on model fitting
and controlling.
The package also provides functionality for high level graphics of estimated effects, see function
plot.bayesx
, plot2d
, plot3d
, plotblock
,
plotmap
, plotsamples
and colorlegend
.
More standard extractor functions and methods for the fitted model objects may be applied, e.g.,
see function summary.bayesx
, fitted.bayesx
,
residuals.bayesx
, samples
, plot.bayesx
, as well as
AIC
, BIC
etc., please see the examples of the help
sites. Predictions for new data based on refitting with weights can be obtained by function
predict.bayesx
.
In addition, it is possible to run arbitrary BayesX program files using function
run.bayesx
. BayesX output files that are stored in a directory may
be read into R calling function read.bayesx.output
.
Nikolaus Umlauf, Thomas Kneib, Stefan Lang, Achim Zeileis.
## to see the package demos
demo(package = "R2BayesX")
Run the code above in your browser using DataLab