DESCRIPTION

@@ -2,7 +2,7 @@ Package: bayesmeta
 Type: Package
 Title: Bayesian Random-Effects Meta-Analysis and Meta-Regression
 Version: 3.1
-Date: 2022-03-23
+Date: 2022-09-14
 Authors@R: c(person(given="Christian", family="Roever", role=c("aut","cre"), 
                     email="christian.roever@med.uni-goettingen.de",
                     comment=c(ORCID="0000-0002-6911-698X")),
@@ -12,7 +12,7 @@ Authors@R: c(person(given="Christian", family="Roever", role=c("aut","cre"),
 Maintainer: Christian Roever <christian.roever@med.uni-goettingen.de>
 Depends: forestplot (>= 2.0.0), metafor (>= 2.0-0), mvtnorm (>= 1.1-1), numDeriv (>= 2016.8-1.1)
 Suggests: compute.es, knitr, rmarkdown, R.rsp
-Description: A collection of functions allowing to derive the posterior distribution of the two parameters in a random-effects meta-analysis, and providing functionality to evaluate joint and marginal posterior probability distributions, predictive distributions, shrinkage effects, posterior predictive p-values, etc.; For more details, see also Roever C (2020) <doi:10.18637/jss.v093.i06>.
+Description: A collection of functions allowing to derive the posterior distribution of the model parameters in random-effects meta-analysis or meta-regression, and providing functionality to evaluate joint and marginal posterior probability distributions, predictive distributions, shrinkage effects, posterior predictive p-values, etc.; For more details, see also Roever C (2020) <doi:10.18637/jss.v093.i06>, or Roever C and Friede T (2022) <arXiv:2209.06004>.
 License: GPL (>=2)
 URL: https://gitlab.gwdg.de/croever/bayesmeta
 VignetteBuilder: knitr, R.rsp

R/bayesmeta.R

@@ -792,7 +792,7 @@ bayesmeta.default <- function(y, sigma, labels=names(y),
                                    c("tau","mu","theta")))
   expectation <- try(integrate(function(x)return(dposterior(x)*x), lower=0, upper=Inf,
                                rel.tol=rel.tol.integrate, abs.tol=abs.tol.integrate)$value, silent=TRUE)
-  if (class(expectation)=="try-error") {
+  if (inherits(expectation, "try-error")) {
     expectation <- NA
     variance <- NA
   }
@@ -800,7 +800,7 @@ bayesmeta.default <- function(y, sigma, labels=names(y),
     variance <- try(integrate(function(x)return(dposterior(x)*(x-expectation)^2), lower=0, upper=Inf,
                               rel.tol=rel.tol.integrate, abs.tol=abs.tol.integrate)$value,
                     silent=TRUE)
-    if (class(variance)=="try-error")
+    if (inherits(variance, "try-error"))
       variance <- NA
   }
   sumstats[c("mean","sd"),"tau"] <- c(expectation, sqrt(variance))

R/bmr.R

@@ -63,9 +63,15 @@ beta.convert <- function(beta, which.beta, d, betanames)
   if (missing(which.beta)) { # default "which.beta"
     betaIdx <- 1:d
   } else {  # try to make sense of "which.beta" argument; general sanity checks:
+    #if (!is.vector(which.beta)
+    #    || (!is.element(class(which.beta),
+    #                    c("numeric", "integer", "logical", "character"))
+    #        | (length(which.beta) > d))) {
+    #  warning("Cannot make sense of \"which.beta\" argument (1).")
+    #}
     if (!is.vector(which.beta)
-        || (!is.element(class(which.beta),
-                        c("numeric", "integer", "logical", "character"))
+        || (!inherits(which.beta,
+                      c("numeric", "integer", "logical", "character"))
             | (length(which.beta) > d))) {
       warning("Cannot make sense of \"which.beta\" argument (1).")
     }
@@ -430,13 +436,23 @@ bmr.default <- function(y, sigma, labels = names(y),
               missing(beta) || (is.matrix(beta) | is.data.frame(beta) | is.vector(beta)))
     if (!missing(beta)) {
       if (!missing(which.beta)) { # general sanity checks
+        #if (!is.vector(which.beta)
+        #    || (!is.element(class(which.beta),
+        #                    c("numeric", "integer", "logical", "character"))
+        #        | ((class(which.beta)=="logical")
+        #            && ((length(which.beta) != d) | (sum(which.beta)!=1)))
+        #        | ((is.element(class(which.beta),
+        #                       c("numeric", "integer", "character"))
+        #            & (length(which.beta) != 1))))) {
+        #  warning("Cannot make sense of \"which.beta\" argument (1).")
+        #}
         if (!is.vector(which.beta)
-            || (!is.element(class(which.beta),
-                            c("numeric", "integer", "logical", "character"))
-                | ((class(which.beta)=="logical")
+            || (!inherits(which.beta,
+                          c("numeric", "integer", "logical", "character"))
+                | (inherits(which.beta, "logical")
                     && ((length(which.beta) != d) | (sum(which.beta)!=1)))
-                | ((is.element(class(which.beta),
-                               c("numeric", "integer", "character"))
+                | ((inherits(which.beta,
+                             c("numeric", "integer", "character"))
                     & (length(which.beta) != 1))))) {
           warning("Cannot make sense of \"which.beta\" argument (1).")
         }
@@ -517,13 +533,23 @@ bmr.default <- function(y, sigma, labels = names(y),
               missing(beta.p) || (is.matrix(beta.p) | is.data.frame(beta.p) | is.vector(beta.p)))
     if (!missing(beta.p)) {
       if (!missing(which.beta)) { # general sanity checks
+        #if (!is.vector(which.beta)
+        #    || (!is.element(class(which.beta),
+        #                    c("numeric", "integer", "logical", "character"))
+        #        | ((class(which.beta)=="logical")
+        #            & ((length(which.beta) != d) | (sum(which.beta) != 1)))
+        #        | ((is.element(class(which.beta),
+        #                       c("numeric", "integer", "character"))
+        #            & (length(which.beta) != 1))))) {
+        #  warning("Cannot make sense of \"which.beta\" argument (1).")
+        #}
         if (!is.vector(which.beta)
-            || (!is.element(class(which.beta),
-                            c("numeric", "integer", "logical", "character"))
-                | ((class(which.beta)=="logical")
-                    & ((length(which.beta) != d) | (sum(which.beta) != 1)))
-                | ((is.element(class(which.beta),
-                               c("numeric", "integer", "character"))
+            || (!inherits(which.beta,
+                          c("numeric", "integer", "logical", "character"))
+                | (inherits(which.beta,"logical")
+                   & ((length(which.beta) != d) | (sum(which.beta) != 1)))
+                | ((inherits(which.beta,
+                             c("numeric", "integer", "character"))
                     & (length(which.beta) != 1))))) {
           warning("Cannot make sense of \"which.beta\" argument (1).")
         }
@@ -1248,7 +1274,7 @@ bmr.default <- function(y, sigma, labels = names(y),
                                lower=0, upper=Inf,
                                rel.tol=rel.tol.integrate,
                                abs.tol=abs.tol.integrate)$value, silent=TRUE)
-  if (class(expectation)=="try-error") {
+  if (inherits(expectation,"try-error")) {
     expectation <- NA
     variance <- NA
   } else {
@@ -1256,7 +1282,7 @@ bmr.default <- function(y, sigma, labels = names(y),
                               lower=0, upper=Inf,
                               rel.tol=rel.tol.integrate,
                               abs.tol=abs.tol.integrate)$value, silent=TRUE)
-    if (class(variance)=="try-error")
+    if (inherits(variance, "try-error"))
       variance <- NA
   }
   sumstats[c("mean","sd"), "tau"] <- c(expectation, sqrt(variance))
@@ -1583,7 +1609,7 @@ summary.bmr <- function(object, X.mean, X.prediction, ...)
 {
   ############################################################
   # default treatment of "X.mean" argument:
-  if (missing(X.mean) || (is.null(X.mean) || is.na(X.mean))) {
+  if (missing(X.mean) || (is.null(X.mean) || all(is.na(X.mean)))) {
       X.mean <- NULL
       meanNumber <- 0
   } else {
@@ -1597,7 +1623,7 @@ summary.bmr <- function(object, X.mean, X.prediction, ...)
     meanNumber <- nrow(X.mean)
   }
   # default treatment of "X.prediction" argument:
-  if (missing(X.prediction) || (is.null(X.prediction) || is.na(X.prediction))) {
+  if (missing(X.prediction) || (is.null(X.prediction) || all(is.na(X.prediction)))) {
       X.prediction <- NULL
       predNumber <- 0
   } else {
@@ -1672,7 +1698,7 @@ summary.bmr <- function(object, X.mean, X.prediction, ...)
         }
       }
       maxi <- optimize(function(b){return(object$dpredict(theta=b, x=X.prediction[i,], mean=FALSE))},
-                       lower=summary.mean[,"95% lower"], upper=summary.mean[,"95% upper"],
+                       lower=summary.pred[,"95% lower"], upper=summary.pred[,"95% upper"],
                        maximum=TRUE)
       summary.pred[i,"mode"] <- maxi$maximum
     }

man/BucherEtAl1997.Rd

@@ -41,7 +41,12 @@
   \emph{Journal of Clinical Epidemiology}, \bold{50}(6):683-691, 1997.
   \doi{10.1016/S0895-4356(97)00049-8}.
 }
-\references{J.P.T. Higgins, J. Thomas, J. Chandler, M. Cumpston, T. Li,
+\references{
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+
+  J.P.T. Higgins, J. Thomas, J. Chandler, M. Cumpston, T. Li,
   M.J. Page, V.A. Welch (eds.).
   \emph{Cochrane handbook for systematic reviews of interventions}.
   Wiley and Sons, 2nd edition, 2019.

man/CrinsEtAl2014.Rd

@@ -46,7 +46,17 @@
 \seealso{
   \code{\link{GoralczykEtAl2011}}.
 }
-\references{T.G. Heffron et al.
+\references{
+  C. Roever.
+  Bayesian random-effects meta-analysis using the bayesmeta R package.
+  \emph{Journal of Statistical Software}, \bold{93}(6):1-51, 2020.
+  \doi{10.18637/jss.v093.i06}.
+  
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+
+  T.G. Heffron et al.
   Pediatric liver transplantation with daclizumab induction therapy.
   \emph{Transplantation}, \bold{75}(12):2040-2043, 2003.
   \doi{10.1097/01.TP.0000065740.69296.DA}.

man/NicholasEtAl2019.Rd

@@ -31,6 +31,10 @@
   \emph{Multiple Sclerosis Journal}, \bold{25}(11):1462-1471, 2019.
   \doi{10.1177/1352458518794063}.
 }
+\references{C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+}
 \examples{
 # load data:
 data("NicholasEtAl2019")

man/RobergeEtAl2017.Rd

@@ -40,6 +40,10 @@
   \bold{216}(2):110-120, 2017.
   \doi{10.1016/j.ajog.2016.09.076}.
 }
+\references{C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+}
 \seealso{
   \code{\link{bmr}}, \code{\link[metafor]{escalc}},
   \code{\link[stats]{model.matrix}}.

man/SchmidliEtAl2017.Rd

@@ -19,7 +19,8 @@
     age-related macular degeneration (AMD)} was considered. Trial
   design, and in particular considerations regarding the required sample
   size, hinge on the expected amount of variability in the primary
-  endpoint (here: \emph{visual acuity}).
+  endpoint (here: \emph{visual acuity}, which is measured on a scale
+  ranging from 0 to 100 via an eye test chart).
   
   Historical data from six previous trials are available (Szabo \emph{et
   al.}; 2015), each trial providing an estimate \eqn{\hat{s}_i}{s[i]} of
@@ -30,7 +31,7 @@
   \deqn{y_i=\log(\hat{s}_i) \quad \mbox{and} \quad 
   \sigma_i=\sqrt{\frac{1}{2\,\nu_i}}}{y[i] = log(s[i])  and  sigma[i] = sqrt(1/(2*nu[i]))}
   
-  The \emph{unit information standard deviation} for a logarithmic
+  The \emph{unit information standard deviation (UISD)} for a logarithmic
   standard deviation then is at approximately
   \eqn{\frac{1}{\sqrt{2}}}{2^-0.5}. 
 }

man/bayesmeta-package.Rd

@@ -5,27 +5,28 @@
   Bayesian Random-Effects Meta-Analysis and Meta-Regression
 }
 \description{
-  Description: A collection of functions allowing to derive the
-  posterior distribution of the two parameters in a random-effects
-  meta-analysis, and providing functionality to evaluate joint and
+  A collection of functions allowing to derive the posterior
+  distribution of the model parameters in random-effects meta-analysis
+  or meta-regression, and providing functionality to evaluate joint and
   marginal posterior probability distributions, predictive
-  distributions, shrinkage effects, posterior predictive p-values,
-  etc.
+  distributions, shrinkage effects, posterior predictive p-values, etc.
 }
 \details{
   \tabular{ll}{
     Package: \tab bayesmeta\cr
     Type:    \tab Package\cr
     Version: \tab 3.1\cr
-    Date:    \tab 2022-03-23\cr
+    Date:    \tab 2022-09-14\cr
     License: \tab GPL (>=2)
   }
   The main functionality is provided by the \code{\link{bayesmeta}()}
-  function. It takes the data (estimates and associated standard
-  errors) and prior information (effect and heterogeneity priors), and
-  returns an object containing functions that allow to derive posterior
-  quantities like joint or marginal densities, quantiles, etc.
-  The \code{\link{bmr}()} function extends the approach to meta-regression.
+  and \code{\link{bmr}()} function. It takes the data (estimates and
+  associated standard errors) and prior information (effect and
+  heterogeneity priors), and returns an object containing functions that
+  allow to derive posterior quantities like joint or marginal densities,
+  quantiles, etc. The \code{\link{bmr}()} function extends the approach
+  to meta-regression by allowing to specify covariables (moderators) in
+  addition.
 }
 \author{
 Christian Roever <christian.roever@med.uni-goettingen.de>
@@ -39,6 +40,10 @@ Christian Roever <christian.roever@med.uni-goettingen.de>
   Bayesian random-effects meta-analysis using the bayesmeta R package.
   \emph{Journal of Statistical Software}, \bold{93}(6):1-51, 2020.
   \doi{10.18637/jss.v093.i06}.
+  
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
 }
 \keyword{ models }
 \keyword{ package }

man/bmr.Rd

@@ -319,6 +319,10 @@
     the \code{bmr} object.}
   }
 \references{
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+
   C. Roever.
   Bayesian random-effects meta-analysis using the bayesmeta R package.
   \emph{Journal of Statistical Software}, \bold{93}(6):1-51, 2020.

man/forestplot.bmr.Rd

@@ -90,6 +90,10 @@
   Christian Roever \email{christian.roever@med.uni-goettingen.de}
 }
 \references{
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+
   C. Roever.
   Bayesian random-effects meta-analysis using the bayesmeta R package.
   \emph{Journal of Statistical Software}, \bold{93}(6):1-51, 2020.

man/forestplot.escalc.Rd

@@ -70,7 +70,8 @@
 \seealso{
   \code{\link[metafor]{escalc}},
   \code{\link[forestplot]{forestplot}},
-  \code{\link{forestplot.bayesmeta}}.
+  \code{\link{forestplot.bayesmeta}},
+  \code{\link{forestplot.bmr}}.
 }
 \examples{
 # load data:

man/summary.bmr.Rd

@@ -53,6 +53,11 @@
 \author{
   Christian Roever \email{christian.roever@med.uni-goettingen.de}
 }
+\references{
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+}
 \examples{
 \dontrun{
 # perform a meta-analysis using binary ("indicator") covariables;

man/traceplot.bayesmeta.Rd

@@ -77,6 +77,10 @@
   \emph{Journal of Statistical Software}, \bold{93}(6):1-51, 2020.
   \doi{10.18637/jss.v093.i06}.
   
+  C. Roever, T. Friede.
+  Using the bayesmeta R package for Bayesian random-effects meta-regression.
+  \href{https://arxiv.org/abs/2209.06004}{arXiv preprint 2209.06004}, 2022.
+
   D.B. Rubin.
   Estimation in parallel randomized experiments.
   \emph{Journal of Educational Statistics}, \bold{6}(4):377-401, 1981.

vignettes/RoeverFriede2022-UsingBayesmetaForMetaRegression.pdf.asis

+%\VignetteIndexEntry{Using the bayesmeta R package for Bayesian random-effects meta-regression}
+%\VignetteEngine{R.rsp::asis}