TCE: Transellptical Correlation Estimation

In smart: Sparse Multivariate Analysis via Rank Transformation

Description

A function to conduct Transellptical Correlation Estimation

Usage

TCE(x, method, nlambda = NULL, lambda.min.ratio = NULL, lambda = NULL, verbose = TRUE)

Arguments

x	The n by d data matrix or d by d covariance matrix from the input
method	Method to be used to estimating the correlation matrix with 5 options: pearson, ns, npn, spearman and kendall. kendall as default.
nlambda	The number of regularization/thresholding paramters. The default value is 20.
lambda.min.ratio	The largest sparsity level for the estimated graphs. The default valye is 0.05.
lambda	A sequence of positive numbers for conducting thresholding.
verbose	If verbose = FALSE, tracing information printing is disabled. The default value is TRUE.

Details

The correlation graph is estimated by correlation cut-off based on the given thresholding level.

Value

An object with S3 class "TCE" is returned:

cov.input	An indicator of the sample covariance.
path	A list of k by k adjacency matrices of estimated graphs as a graph path corresponding to lambda.
sparsity	The sparsity levels of the graph path.
method	The method used in the correlation graph estimation stage.
lambda	The sequence of thresholding parameters used.

Author(s)

Fang Han, Han Liu
Maintainer: Fang Han<[e-mail]>

References

1.Han Liu, Fang Han, Ming Yuan, John Lafferty, Larry Wasserman. High dimensional semiparametric gaussian copula graphical models. Annals of Statistics, to apear.
2.Tuo Zhao and Han Liu. HUGE: A Package for High-dimensional Undirected Graph Estimation. Technical Report, Carnegie Mellon University, 2010

Examples

require(huge)
L = huge.generator(n = 200, d = 80, graph = "hub")
out = TCE(L$data,method="kendall")
out
plot(out)

Example output

Loading required package: huge
Loading required package: Matrix
Loading required package: lattice
Loading required package: igraph

Attaching package: 'igraph'

The following objects are masked from 'package:stats':

    decompose, spectrum

The following object is masked from 'package:base':

    union

Loading required package: MASS
Generating data from the multivariate normal distribution with the hub graph structure....done.
Conducting nonparanormal (npn) transformation via kendall....done.
Conducting the transelliptical correlation estimation:5% 
Conducting the transelliptical correlation estimation:10% 
Conducting the transelliptical correlation estimation:15% 
Conducting the transelliptical correlation estimation:20% 
Conducting the transelliptical correlation estimation:25% 
Conducting the transelliptical correlation estimation:30% 
Conducting the transelliptical correlation estimation:35% 
Conducting the transelliptical correlation estimation:40% 
Conducting the transelliptical correlation estimation:45% 
Conducting the transelliptical correlation estimation:50% 
Conducting the transelliptical correlation estimation:55% 
Conducting the transelliptical correlation estimation:60% 
Conducting the transelliptical correlation estimation:65% 
Conducting the transelliptical correlation estimation:70% 
Conducting the transelliptical correlation estimation:75% 
Conducting the transelliptical correlation estimation:80% 
Conducting the transelliptical correlation estimation:85% 
Conducting the transelliptical correlation estimation:90% 
Conducting the transelliptical correlation estimation:95% 
Conducting the transelliptical correlation estimation:100% 
done 
Transelliptical Correlation Estimation via Thresholding 
Method: kendall 
Input: The Data Matrix
Path length: 20 
Sparsity level: 0.0003164557 -----> 0.05 

smart documentation built on May 29, 2017, 8:58 p.m.