version 1.11

DESCRIPTION

@@ -1,14 +1,14 @@
 Package: glasso
 Title: Graphical Lasso: Estimation of Gaussian Graphical Models
-Version: 1.10
+Version: 1.11
 Author: Jerome Friedman, Trevor Hastie and Rob Tibshirani
 Description: Estimation of a sparse inverse covariance matrix using a lasso (L1)
              penalty. Facilities are provided for estimates along a path of values
 	     for the regularization parameter.
 Maintainer: Rob Tibshirani <tibs@stat.stanford.edu>
 License: GPL-2
 URL: http://www-stat.stanford.edu/~tibs/glasso
-Packaged: 2018-07-12 23:45:19 UTC; naras
+Packaged: 2019-09-24 21:37:23 UTC; robtibshirani
 Repository: CRAN
-Date/Publication: 2018-07-13 04:40:06 UTC
+Date/Publication: 2019-10-01 18:40:07 UTC
 NeedsCompilation: yes

MD5

@@ -1,8 +1,8 @@
-9e100048db5c6fbde74725c20d32d8ab *DESCRIPTION
+221127194e4b26a07e13985c40925a45 *DESCRIPTION
 a98919657d182a30729381c29aa13061 *NAMESPACE
-9b2f2da42cd1d66c4e3cd38427a0bf7a *R/glasso.R
+4d3c6b4c4db738159f953e00df2394f7 *R/glasso.R
 c1a3c7a42d3c130ba02f4827f056ac9a *R/glassopath.R
-a92a091ed68d9cf0d5ee056b4ead20fb *man/glasso.Rd
+eda53f6f766738b1f564751d5d299160 *man/glasso.Rd
 e624eb3357a4d5c9dd0f2c5c734c7116 *man/glassopath.Rd
 eec797dc32427aee0efb399871b455b4 *src/glasso.f
 22909b3dcaec3bcf1cd20d2fe6134966 *src/glasso_init.c

R/glasso.R

@@ -1,8 +1,8 @@
-glasso=function(s, rho, zero=NULL, thr=1.0e-4,maxit=1e4,  approx=FALSE, penalize.diagonal=TRUE,start=c("cold","warm"), w.init=NULL,wi.init=NULL, trace=FALSE){
-n=nrow(s)
+glasso=function(s, rho,nobs=NULL, zero=NULL, thr=1.0e-4,maxit=1e4,  approx=FALSE, penalize.diagonal=TRUE,start=c("cold","warm"), w.init=NULL,wi.init=NULL, trace=FALSE){
 # on return, cflag=1 means the procedure did not converge, =0 means  it did
 
 BIG=10e9
+n=nrow(s)
 
 if(!is.matrix(rho) & length(rho)!=1 & length(rho)!=nrow(s))
    {stop("Wrong number of elements in rho")}
@@ -86,5 +86,7 @@ val= -log(d)+sum(diag(s%*%Sigmahati))+sum(abs(rho*temp))
 return(val)
 }
 crit=critfun(xx,s,rho,penalize.diagonal)
-return(list(w=ww,wi=xx,loglik=-(n/2)*crit,errflag=junk$ierr,approx=approx, del=junk$del, niter=junk$niter))
+loglik=NA
+if(!is.null(nobs)) loglik=-(nobs/2)*crit
+return(list(w=ww,wi=xx,loglik=loglik,errflag=junk$ierr,approx=approx, del=junk$del, niter=junk$niter))
 }

man/glasso.Rd

@@ -5,13 +5,15 @@
 Estimates a sparse inverse covariance matrix using a lasso (L1) penalty
 }
 \usage{
-glasso(s, rho, zero=NULL, thr=1.0e-4, maxit=1e4,  approx=FALSE, 
+glasso(s, rho, nobs=NULL, zero=NULL, thr=1.0e-4, maxit=1e4,  approx=FALSE, 
 penalize.diagonal=TRUE, start=c("cold","warm"), 
 w.init=NULL,wi.init=NULL, trace=FALSE)
 }
 \arguments{
 \item{s}{Covariance matrix:p by p matrix (symmetric)}
 \item{rho}{(Non-negative) regularization parameter for lasso. rho=0 means no regularization. Can be a scalar (usual) or a symmetric p by p matrix, or a vector of length p. In the latter case, the penalty matrix has jkth element sqrt(rho[j]*rho[k]).}
+\item{nobs}{Number of observations used in computation of the  covariance matrix s. This quantity is need to compute the value of  log-likelihood.
+If not specified, loglik will be returned as NA.}
 \item{zero}{(Optional) indices of entries of inverse covariance to be constrained to be zero. The input should be a matrix with two columns, each row indicating
 the indices of elements to be constrained to be zero. The solution must be symmetric, so you need only  specify one of (j,k) and (k,j). An entry in the zero matrix
 overrides any entry in the rho matrix for a given element.}