bic: Lasso Bayesian Information Criterion

In Marga8/HDGCvar: Granger Causality Testing in High Dimensional Vector Autoregressive Models

Lasso Bayesian Information Criterion

Description

Do BIC on lasso output. Needs degrees of freedom coming from glmnet. Bound on lambda is optional, imposed as 0.5T by default.

Usage

bic(y, x, b, df, bound = 0.5 * length(y))

Arguments

y	vector of dependent variable
x	vector or matrix of regressors
b	lasso coefficients (taken from glmnet)
df	lasso degrees of freedom (taken from glmnet)
bound	upper bound on tuning parameter lambda: i.e. impose lasso to stop the selection after it selects 0.5T parameters.

Value

the estimated parameters with BIC

Marga8/HDGCvar documentation built on May 25, 2024, 11:12 a.m.