Learn R Programming

hierband (version 1.0)

Convex Banding of the Covariance Matrix

Description

Implementation of the convex banding procedure (using a hierarchical group lasso penalty) for covariance estimation that is introduced in Bien, Bunea, Xiao (2015) Convex Banding of the Covariance Matrix. Accepted for publication in JASA.

Copy Link

Version

Install

install.packages('hierband')

Monthly Downloads

16

Version

1.0

License

GPL-3

Maintainer

Jacob Bien

Last Published

June 12th, 2015

Functions in hierband (1.0)

hierband

Solves main optimization problem for fixed lambda value
ma

Covariance of an equal-weighted moving-average process
subdiag.thresh

Performs a single pass of BCD on a matrix R.
hierband-package

Convex banding of the covariance matrix using
hierband.path

Solves main optimization problem over a grid of lambda values
gpband

Groupwise soft-thresholds subdiagonals by lam * w
formw

Form the "general weights" matrix
subdiagonal.l2norms

Compute the L2 norm of each subdiagonal of a symmetric matrix R.
lam.max.hierband

Computes the smallest lambda such that P=0.
MakeFolds

Make folds for cross validation
hierband.cv

Performs nfolds-cross validation
banded

Generates a banded covariance matrix and matrix squareroot sig: value of kth band (starting with diagonal) size of band is length(sig)