R help - Jan 2011 - Multivariate autoregressive models with lasso penalization

I wish to estimate sparse causal networks from simulated time series data.
Although there's some discussion about this problem in the literature (at
least a few authors have used lasso and l(1,2) regularization to enforce
sparsity in multivariate autoregressive models, e.g.,
http://user.cs.tu-berlin.de/~nkraemer/papers/grplasso_causality.pdf), I
can't find any R packages with these capabilities.

Has anyone in the R community experimented with such or put code out for
this problem?

Many thanks.
John

-- 
John M. Drake, Ph.D.
Associate Professor
University of Georgia
Odum School of Ecology
Athens, GA 30602-2202

phone:  706.583.5539
fax:       706.542.4819
email:   jdrake@uga.edu
skype:  john.drake.uga
web:     http://dragonfly.ecology.uga.edu/drakelab

	[[alternative HTML version deleted]]