Description Usage Arguments Value
BayesGLASSO
Provides a regularized precision matrix estimate using
a Bayesian GLASSO.
1 2 | BayesGLASSO(data, lambda_shape = 1, lambda_rate = 0.01, nBurnin = 10000,
nIter = 10000, verbose = TRUE, lambda)
|
data |
A data matrix with rows representinh participants and columns representing variables/nodes. |
lambda_shape |
The rate parameter for the hyperprior of lambda, the parameter that determines the amount of shirnkage. |
lambda_rate |
The scale parameter for the hyperprior of lambda, the parameter that determines the amount of shirnkage. |
nIter |
The number of iterations for the Gibbs-sampler. |
verbose |
If 'TRUE'displays a progress bar. |
lambda |
The shrinkage parameter. This value is estimated by default. If given here lambda won't be update as part of the Gibbs-sampler. |
nBurning |
The number of burn-in iterations for the Gibbs-sampler. |
The function returns the following output:
Omega |
A dataframe containing the estimated precision matrix for each itteration of the Gibbs-sampler. |
pcor |
A dataframe containing the estimated partial correlation matrix for each itteration of the Gibbs-sampler. |
lambda_sq |
A dataframe containing the estimates for the local hyper parameter lambda for each itteration of the Gibbs-sampler. |
tau_sq |
A dataframe containing the estimates for the global hyper parameter tau for each itteration of the Gibbs-sampler. |
Strength |
A dataframe containing the estimated strength centrality for each node for each itteration of the Gibbs-sampler. |
Closeness |
A dataframe containing the estimated closeness centrality for each node for each itteration of the Gibbs-sampler. |
Betweenness |
A dataframe containing the estimated betweenness centrality for each node for each itteration of the Gibbs-sampler. |
optwi |
The point estimate for the precision matrix obtained by taking the mode of the posterior distribution. |
optpcor |
The point estimate for the partial correlation matrix obtained by taking the mode of the posterior distribution. |
CredInt |
The 95% Credibility interval for the elements of the precision matrix. |
optStrength |
The point estimate for strength centrality of each node based on the point estimate of the partial correlation matrix. |
CIStrength |
The 95% Credibility Interval for the strength centrality of each node. |
optCloseness |
The point estimate for closeness centrality of each node based on the point estimate of the partial correlation matrix. |
CICloseness |
The 95% Credibility Interval for the closeness centrality of each node. |
optBetween |
The point estimate for betweenness centrality of each node based on the point estimate of the partial correlation matrix. |
CIBetween |
The 95% Credibility Interval for the betweenness centrality of each node. |
The following output is only returned in case of missing data.
Missing |
Indicator for which observations where missing. |
ImputedValues |
The imputed values for each missing datapoint for each itteration of the Gibbs-sampler. |
CompleteData |
The original dataset with missing values replaced by the mean imputed value for each missing data point. |
OriginalData |
The original dataset with missing values. |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.