spam.Rmd
SIGNIFICANT USERVISIBLE CHANGES
INTERNAL CHANGES
INTERNAL CHANGES
INTERNAL CHANGES
BUG FIXES
diff git a/public/reference/Oral.html b/public/reference/Oral.html index 47fb666b4e0f7acf2e6c598dd93620b56530b066..8a336cf5bcc91e0277fb6d066525461563ed2374 100644  a/public/reference/Oral.html +++ b/public/reference/Oral.html @@ 143,8 +143,7 @@ adjacency graph.The data is available from the package INLA
 distributed from www.rinla.org or from
 http://www.math.ntnu.no/~hrue/GMRFbook/oral.txt
chol
performs a Cholesky
decomposition of a symmetric positive definite sparse matrix x
+decomposition of a symmetric positive definite sparse matrix x
of class spam
.
summary
call, see ‘Examples&
backsolve
to solve a system of linear equations.Notice that the Cholesky factorization of the package SparseM
is also
based on the algorithm of Ng and Peyton (1993). Whereas the Cholesky
routine of the package Matrix
are based on
+routine of the package Matrix
are based on
CHOLMOD
by Timothy A. Davis (C
code).
TRUE
).
det.spam
, solve.spam
,
+
TRUE
).
R < as.spam(cholS)
mvsample < ( array(rnorm(N*n),c(N,n)) %*% R)[,iord]
# It is often better to order the sample than the matrix
# R itself.
+# R itself.
# 'mvsample' is of class 'spam'. We need to transform it to a
# regular matrix, as there is no method 'var' for 'spam' (should there?).
diff git a/public/reference/rmvnorm.const.html b/public/reference/rmvnorm.const.html
index 9e4226ff81e79171f84fc6201cd56e117d3b2b85..e79dfcf2b052a3fff4ddb2b31227887e2adc149d 100644
 a/public/reference/rmvnorm.const.html
+++ b/public/reference/rmvnorm.const.html
@@ 124,12 +124,12 @@
rmvnorm.const(n, mu = rep(0, nrow(Sigma)), Sigma, Rstruct = NULL,  A = array(1, c(1,nrow(Sigma))), a=0, U=NULL, ...) rmvnorm.prec.const(n, mu = rep(0, nrow(Q)), Q, Rstruct = NULL,  A = array(1, c(1,nrow(Q))), a=0, U=NULL, ...) +rmvnorm.const(n, mu = rep.int(0, dim(Sigma)[1]), Sigma, Rstruct = NULL, + A = array(1, c(1,dim(Sigma)[1])), a=0, U=NULL, ...) +rmvnorm.prec.const(n, mu = rep.int(0, dim(Q)[1]), Q, Rstruct = NULL, + A = array(1, c(1,dim(Q)[1])), a=0, U=NULL, ...) rmvnorm.canonical.const(n, b, Q, Rstruct = NULL,  A = array(1, c(1,nrow(Q))), a=0, U=NULL, ...)+ A = array(1, c(1,dim(Q)[1])), a=0, U=NULL, ...)
The functions rmvnorm.prec
and rmvnorm.canonical
 do not requrie sparse precision matrices.
+ do not requrie sparse precision matrices.
For rmvnorm.spam
, the differences between regular and sparse
covariance matrices are too significant to be implemented here.
Often (e.g., in a Gibbs sampler setting), the sparsity structure of
diff git a/public/reference/rmvnorm.html b/public/reference/rmvnorm.html
index 88466477cdea686318f8289fcec6537439dfe43b..3a5577499eb26c148dc1db748e3c41428c95660f 100644
 a/public/reference/rmvnorm.html
+++ b/public/reference/rmvnorm.html
@@ 124,8 +124,8 @@

rmvnorm.spam(n,mu=rep(0, nrow(Sigma)), Sigma, Rstruct=NULL, ...) rmvnorm.prec(n,mu=rep(0, nrow(Q)), Q, Rstruct=NULL, ...) +is<sample(ln,nz) js<sample(ln,nz) system.time(for (iin1:nz) smat[is[i], js[i]] <i)rmvnorm.spam(n,mu=rep.int(0, dim(Sigma)[1]), Sigma, Rstruct=NULL, ...) +rmvnorm.prec(n,mu=rep.int(0, dim(Q)[1]), Q, Rstruct=NULL, ...) rmvnorm.canonical(n, b, Q, Rstruct=NULL, ...)Arguments
@@ 164,7 +164,7 @@Details
The functions
rmvnorm.prec
andrmvnorm.canonical
 do not require sparse precision matrices. + do not require sparse precision matrices. Forrmvnorm.spam
, the differences between regular and sparse covariance matrices are too significant to be implemented here.
Often (e.g., in a Gibbs sampler setting), the sparsity structure of @@ 192,14 +192,14 @@ Sigmainv < as.spam( Sigmainv, eps=1e4) Sigma < solve( Sigmainv) # for verification +Sigma < solve( Sigmainv) # for verification iidsample < array(rnorm(N*n),c(n,N)) mvsample < backsolve( chol(Sigmainv), iidsample) norm( var(t(mvsample))  Sigma, type="m")#> [1] 0.1326448# compare with: mvsample < backsolve( chol(as.matrix( Sigmainv)), iidsample, n)  #### ,n as patch + #### ,n as patch norm( var(t(mvsample))  Sigma, type="m")#> [1] 0.1326447diff git a/public/reference/spam.creation.html b/public/reference/spam.creation.html index b38a6cb22c2b11066d2c584fbe4ece6f8e113300..ae07a549054676b7fd643a0b6590f19746e5ab0f 100644  a/public/reference/spam.creation.html +++ b/public/reference/spam.creation.html @@ 212,7 +212,7 @@ is.spam(x)
+spam.version$version.string#> [1] "Spam version 2.21 (20181011)"
spam.version$version.string#> [1] "Spam version 2.21 (20181220)"