Below, you find development versions of a couple of R packages I am (co-)authoring. I strongly recommend using stable versions (available on CRAN) of these packages for everyday's usage. However, if you really need a brand new feature of one of these packages which is not present in the stable version yet, you may consider installing the development versions provided here. Please notice that these pre-release versions may (or will) be buggy, and subject to frequent changes.

pcalg

This package contains several functions for causal structure learning and causal inference using graphical models. The main algorithms for causal structure learning are PC and GES (for observational data without hidden variables), FCI and RFCI (for observational data with hidden variables), and GIES (for a mix of observational and interventional data without hidden variables). For causal inference the IDA algorithm and the generalized backdoor criterion is implemented.

Differences of the development to the stable version: the development version contains an early implementation of ARGES (adaptively restricted greedy equivalence search), a variant of GES which, together with an estimator of the conditional independence graph (CIG), consistently estimates a CPDAG.

The recommended stable version of the package is available on CRAN.

The development version can be installed as follows:

• Source package:

  install.packages("pcalg", repos = "http://www.huschhus.ch/R", type = "source")

• Windows binary package:

  install.packages("pcalg", repos = "http://www.huschhus.ch/R", type = "win.binary")

sfsmisc

This package contains useful utilities ('goodies') from Seminar fuer Statistik ETH Zurich, quite a few related to graphics; many ported from S-plus times.

No development version at the moment; please consider the stable version on CRAN.