Release to CRAN

.Rbuildignore

@@ -6,3 +6,4 @@ NAMESPACE_RG2
 notebooks
 reports
 extra_tests
+releases

DESCRIPTION

@@ -36,7 +36,7 @@ Suggests:
     modeest,
     testthat (>= 3.0.0),
     knitr,
-    rmarkdown,
+    markdown,
     spelling
 Config/testthat/edition: 3
 VignetteBuilder: knitr

README → INSTALL

@@ -1,102 +1,107 @@
-######################################################################################################
-###
-###  runjags package
-###  Matthew Denwood, 2nd October 2015
-###
-###  Installation from source - help and tips
-###
-###  The instructions below are modified from the rjags package
-###  Original file is Copyright (C) Martyn Plummer, licensed under GPL-2
-###
-######################################################################################################
-
-
-The runjags package itself is entirely written in R, but it does contain
-a JAGS module that provides additional distributions that must be
-compiled for your system against a functional JAGS installation.  See
-the platform specific instructions below for how to do this.
-
-If you are unable to compile the C++ code and don't require the distributions
-provided in the JAGS extension module, see the 'Compile-free' instructions 
-at the end of this document.
-
-
-
-Windows
--------
-
-At compile time, there is no way for Windows to guess where the JAGS
-library is installed, so you must supply the location yourself using
-the make variable JAGS_ROOT, defining it in the file HOME/.R/Makevars.win
-like so:
-
-JAGS_ROOT=c:\Program Files\JAGS\JAGS-3.3.0
-
-The configure script also needs to know the version of JAGS being used,
-and will attempt to guess this from the JAGS_ROOT path.  If a non-
-standard install path has been used you will need to provide the correct
-JAGS major-version number like so:
-
-JAGS_MAJOR_VERSION=3
-
-The header files for the module itself should be able to detect the
-version of JAGS installed on your system, but if you encounter compiler
-errors it may help to specify JAGS_MAJOR_VERSION even if JAGS is 
-installed in a standard directory - e.g. JAGS 4-beta requires:
-
-JAGS_MAJOR_VERSION=4
-
-
-
-Unix (including Mac OS X)
--------------------------
-
-The configure script will try to guess the location of the JAGS
-library installation. It will look in /<prefix>/<lib>, where
-
-- <prefix> is the determined from the location of the jags
-  script that launches the command line interface.
-
-  For example, if this is "/usr/local/bin/jags", then prefix will
-  be "/usr/local"
-
-- <lib> is derived from the value of "LIBnn", returned by 
-  "R CMD config LIBnn".
-
-  This is particularly important on 64-bit unix, where R is 
-  installed by default in /usr/local/lib64 and LIBnn is "lib64"
-
-The configure script will also look for JAGS headers in
-/<prefix>/include/JAGS
-
-If the configure script fails to find the JAGS headers or library,
-you can set the environment variables JAGS_INCLUDE and JAGS_LIB
-to point to the correct directory.
-
-The header files for the module itself should be able to detect the
-version of JAGS installed on your system, but a JAGS_MAJOR_VERSION
-environmental variable can be set in case of problems which may be
-encountered with some versions of JAGS - e.g. JAGS 4-beta requires:
-
-export JAGS_MAJOR_VERSION=4
-R CMD install runjags...
-
-
-
-Compile-free
-------------
-
-The standard version of the runjags package includes a JAGS
-extension module providing additional distributions in the Pareto
-family (amongst others).  
-
-If you want to install runjags without the additional JAGS module,
-then you can simply remove the configure(.ac) files and entire src
-directory before running R CMD install runjags...  This does not
-require any C++ compilers, or even for JAGS to be installed on
-your system.
-
-Alternatively there is a version of runjags with the JAGS module 
-removed hosted on the sourceforge page for runjags at:
-http://sourceforge.net/projects/runjags/).
-
+######################################################################################################
+###
+###  Installation instructions for the runjags package
+###  Matthew Denwood, 27th February 2021
+###
+###  The instructions below are modified from the rjags package
+###  Original file is Copyright (C) Martyn Plummer, licensed under GPL-2
+###
+######################################################################################################
+
+
+The runjags package itself is entirely written in R, but it does contain
+a JAGS module that provides additional distributions that must be
+compiled for your system against a functional JAGS installation.  See
+the platform specific instructions below for how to do this.
+
+If you are unable to compile the C++ code and don't require the distributions
+provided in the JAGS extension module, see the 'Compile-free' instructions 
+at the end of this document.
+
+
+
+Installation on Windows
+-----------------------
+
+At compile time, there is no way for Windows to guess where the JAGS
+library is installed, so you must supply the location yourself using
+the make variable JAGS_ROOT, defining it in the file HOME/.R/Makevars.win
+like so:
+
+JAGS_ROOT=c:/Progra~1/JAGS/JAGS-4.0.0
+
+where HOME is the default starting directory for an R session launched
+from the start menu (typically your Documents folder).
+
+The configure script also needs to know the version of JAGS being used,
+and will attempt to guess this from the JAGS_ROOT path.  If a non-
+standard install path has been used you will need to provide the correct
+JAGS major-version number like so:
+
+JAGS_MAJOR_VERSION=4
+
+The header files for the module itself should be able to detect the
+version of JAGS installed on your system, but if you encounter compiler
+errors it may help to specify JAGS_MAJOR_VERSION even if JAGS is 
+installed in a standard directory.
+
+
+
+Unix/Linux: pkg-config
+------------------------
+
+If you have pkg-config installed then the configure script will use it
+to detect the JAGS library and set the compiler and linker flags for
+runjags. You can test this capability in a unix shell by typing
+
+pkg-config --modversion jags
+
+If this does not detect your JAGS installation then it means that the
+file 'jags.pc' is not on the search path of pkg-config. In this case
+you should add it by setting the environment variable PKG_CONFIG_PATH.
+For example if the full path to 'jags.pc' is
+'/usr/local/lib64/pkgconfig/jags.pc', then set
+
+export PKG_CONFIG_PATH=/usr/local/lib64/pkgconfig
+
+before installing runjags.
+
+Recall that most Linux distributions divide software into `user' and
+`developer' versions, so if you are using a pre-built JAGS binary
+package for your Linux distribution then remember to install the
+developer version before attempting to install the runjags package.
+
+The header files for the module itself should be able to detect the
+version of JAGS installed on your system, but a JAGS_MAJOR_VERSION
+environmental variable can be set in case of problems which may be
+encountered with some versions of JAGS - e.g. JAGS 4-beta requires:
+
+export JAGS_MAJOR_VERSION=4
+R CMD install runjags
+
+
+
+Unix/Linux: legacy configuration
+----------------------------------
+
+Legacy configuration is used whenever pkg-config is not installed or
+otherwise fails to detect the JAGS library. Legacy configuration is
+deprecated and will be removed in a future version. If you experience
+problems with installing runjags then it is strongly recommended that
+you try to get pkg-config working. Otherwise, you can refer to the
+legacy installation instructions given within the rjags package.
+
+
+
+Compile-free
+------------
+
+The standard version of the runjags package includes a JAGS
+extension module providing additional distributions in the Pareto
+family (amongst others).  
+
+If you want to install runjags without the additional JAGS module,
+then you can simply remove the configure(.ac) files and entire src
+directory before running R CMD install runjags.  This does not
+require any C++ compilers, or even for JAGS to be installed on
+your system.

NAMESPACE

@@ -63,6 +63,7 @@ write.jagsfile,
 write.JAGSfile,
 template.jags,
 template.JAGS,
+template_huiwalter,
 
 extract.runjags,
 extract,

NAMESPACE_OLD

@@ -63,6 +63,7 @@ write.jagsfile,
 write.JAGSfile,
 template.jags,
 template.JAGS,
+template_huiwalter,
 
 extract.runjags,
 extract,

R/template_huiwalter.R

@@ -2,14 +2,15 @@
 #'
 #' @param testdata the input paired test data, where each column name corresponds to a test result - except possibly "ID" which is ignored, and "Population" indicating a population identifier for that row. Each row must represent test results from the same individual either as logical or a factor with two levels (and where the first level indicates a negative test result). Data may be missing at random (except for Population).
 #' @param outfile the name of the text file to save the model representation
-#' @param covon should covariance terms be activated or omitted?
+#' @param covariance should covariance terms be activated or omitted?
 #' @param se_priors the priors to use for sensitivity parameters (can be adjusted in the model once it is generated)
 #' @param sp_priors the priors to use for specificity parameters (can be adjusted in the model once it is generated)
 #'
 #' @export
-template_huiwalter <- function(testdata, outfile='huiwalter_model.txt', covon=FALSE, se_priors='dbeta(1,1)', sp_priors='dbeta(1,1)'){
+template_huiwalter <- function(testdata, outfile='huiwalter_model.txt', covariance=FALSE, se_priors='dbeta(1,1)', sp_priors='dbeta(1,1)'){
 
 	stopifnot(is.data.frame(testdata))
+	covon <- covariance
 
 	## R code to generate a Hui-Walter model for N tests and P populations, with potential missing data
 
@@ -36,7 +37,7 @@ template_huiwalter <- function(testdata, outfile='huiwalter_model.txt', covon=FA
 	npop <- length(levels(testdata$Population))
 
 	## Initialise the file:
-	cat('## Auto-generated Hui-Walter model created by script version ', version, ' on ', as.character(Sys.Date()), '\n\nmodel{\n\n\t## Observation layer:', sep='', file=outfile, append=FALSE)
+	cat('## Auto-generated Hui-Walter model created by runjags version ', runjagsprivate$runjagsversion, ' on ', as.character(Sys.Date()), '\n\nmodel{\n\n\t## Observation layer:', sep='', file=outfile, append=FALSE)
 
 	## Some variables that are needed in a few places:
 	testcols <- names(testdata)[!names(testdata) %in% c('ID','Population')]