estimate.covariance: Covariance estimation

In sparsebn: Learning Sparse Bayesian Networks from High-Dimensional Data

Description

Methods for inferring (i) Covariance matrices and (ii) Precision matrices for continuous, Gaussian data.

Usage

estimate.covariance(data, ...)

estimate.precision(data, ...)

Arguments

data	data as sparsebnData object.
...	(optional) additional parameters to estimate.dag

Details

For Gaussian data, the precision matrix corresponds to an undirected graphical model for the distribution. This undirected graph can be tied to the corresponding directed graphical model; see Sections 2.1 and 2.2 (equation (6)) of Aragam and Zhou (2015) for more details.

Value

Solution path as a plain list. Each component is a Matrix corresponding to an estimate of the covariance or precision (inverse covariance) matrix for a given value of lambda.

Examples

data(cytometryContinuous)
dat <- sparsebnData(cytometryContinuous$data, type = "c", ivn = cytometryContinuous$ivn)
estimate.covariance(dat) # estimate covariance
estimate.precision(dat)  # estimate precision

sparsebn documentation built on Sept. 13, 2020, 5:10 p.m.