imset.R - a suite of functions for R (version 1.3)
imset.R is a suite of functions in R for integer valued multisets (imsets) of graphical models.
A detailed exposition to the theory of imsets can be found in the book by
- Milan Studeny: Probabilistic Conditional Independence Structures, Springer-Verlag 2005.
A description of the conversion functions implemented in imset.R can be found in the paper by
- M. Studeny and J. Vomlel: Transition between graphical and
algebraic representatives of Bayesian network models, In Proceeding of
the 2nd European Workshop on Probabilistic Graphical Models
(PGM’04), Leiden, the Netherlands. It is also available online.
A conference paper about the R implementation of learning Bayesian
networks using imsets (which is part of the imset.R suite) is available:
- J. Vomlel and M. Studeny. Using imsets for learning Bayesian networks,
In Proceedings of the 10th Czech-Japan Seminar on
Data Analysis and Decision Making under Uncertainty, Liblice, Czech Republic, 2007.
It is also available online.
You can dowload the imset.R file and save it into a working directory of your machine.
To use the imset suite within R you need two packages:
- graph - a package that implements some simple graph handling capabilities that is part of the Bioconductor project and
- ggm - a package from CRAN containing functions for fitting Gaussian Markov models.
To install a package start R and use the Package menu. For both of the two required packages repeat the following procedure:
- First, select the repository - CRAN for ggm and Bioconductor for graph.
- Second, select Install package(s) item in the Package menu and select the required package.
Then, start the R console, change directory to the directory, where the file imset.R is located, and type source("imset.R").
A short manual for the imset.R suite is available here.
An example of an R session and data set for this example are also available.