[R-SIG-Finance] Introducing tawny, a package for filtering correlation matrices via random matrix theory and shrinkage estimation

B. Rowe public_browe at muxspace.com
Mon Apr 6 08:37:51 CEST 2009


I am pleased to announce the initial version of tawny, a portfolio
optimization package in R. Tawny is an R package for studying various
correlation matrix filtering methods applied to asset returns. The
filtering techniques included in the package are random matrix theory
and shrinkage estimation. In addition to the filtering methods, also
included in the package is a simple portfolio optimizer and functions to
measure the effectiveness of the methods, including an implementation of
the Kullback-Leibler divergence function.

More information, code samples, and images can be found at:

The code should be available on CRAN in a while, but in the mean time it
is possible to download it directly via the link above.

To run tawny, you will also need to install a utility package that tawny
depends on:

Pending availability on CRAN, you can download it from the above link.

Both packages are licensed GPL-2. Comments and questions are welcome.

Brian Lee Yung Rowe

More information about the R-SIG-Finance mailing list