The ArvoRe package is a Cost-effectiveness Analysis (CEA) implementation for R oriented
to compute problems that involve simple decision tree models and Markov
models. It offer a graphic user interface (GUI) developed in Tcl/Tk.
This graphical interface simplifies the decision tree creation task and
its manipulation, such as, nodes addition, probability values
definition, node kind configuration, etc.
Graphics,
tables and important summary statistics for the ACE are available
trough of windows, the which ones concentrate them in an unique
location. All graphics and decision tree can be exported to image files
using PNG, JPG and BMP format. Some tables with summary statistics of
an CEA problem in this package can be exported for files in the format
CSV (comma separated values) or TXT (text plain). Other important ArvoRe implementation
is the CEA problems solution that involve Markov chains. These problems
are solved by means of first order Monte Carlo simulation. A model
created to a Cost-effectiveness problem in the ArvoRe can
be saved in a file with extension ARV. The main user preferences
settings - as kind of tree to be exhibited, definitions that should be
also exhibited are conserved in this extension file ARV.
REQUIREMENTS
To run ArvoRe is required R version 2.15.2 or superior.
ArvoRe requires R version 2.15.2 or newer and some packages: abind, gplots, gdata, stringr, tcltk, tcltk2 and triangle. This packages must be installed in your system. These packages can be downloaded from Comprehensive R Archive Network, CRAN http://www.R-project.org .
RUNNING ARVORE
Some
arvoRe functions are not standarized, and return some errors during
'RCMD build' command, to build a regular package. However, you can run
arvoRe in your system folowing some steps (windows plataform):
Install R version 2.15.2, or superior. Direct link http://cran.r-project.org/bin/windows/base/ ;
Download arvoRe binary zip below;
Unzip arvoRe_0.1.XX.zip in "C:\Program Files\R\R-2.15.2\library folder using WinZip (or similar software, Winrar for instance);
Run this script on R-GUI command prompt:
# -----8<----- diretorio <- "C:/Program Files/R/R-2.15.2/library/arvoRe/R" setwd("C:/Program Files/R/R-2.15.2/library/arvoRe/R") listafiles <- list.files(diretorio, pattern = ".R") for (i in listafiles) { print( paste("Processando ",i, sep="") ) source(i) } rm(diretorio) rm(listafiles) # Executa o arvoRe arvore() # -----8<----- |
REMARKS :
arvoRe package try automaticaly install all
dependences. However, some packages, like gdata, some times, show error
in instalation proccess when using R version 2.15.2. If you receive an
error message, delete the "gdata" folder in R library folder
(C:\Program
Files\R\R-2.15.2\library). Then try to run arvoRe again.
# -----8<----- arvore() # -----8<----- |
DEVELOPMENT
ArvoRe package is developed through a SVN's System in R-Forge. Anybody can contribute with improvement and/or development of code. Visit ArvoRe project page.
A more recent and unstable copy of this package can be directly installed in the R using the command
install.packages("arvore",repos="http://R-Forge.R-project.org")
Development versions, the "latest snapshot", can be acquired below.
REFERENCE / HELP
You can read the paper "Implementação de uma interface gráfica para Análise de Custo Efetividade no R" to solve instalation problems and know more about Cost-Effectiveness Analysis.
DOWNLOAD
Binary (Windows)
Source code
EXAMPLE
To
illustrate the arvoRe package, is used an example from the book
"Essentials of Pharmacoeconomics" Karen Rascati. The simulation
include a second order Markov model. The example was modified to
perform a stochastic evaluation. This model try to assess the cost-effectiveness of double antiretroviral therapy (DAT) versus triple antiretroviral therapy (TAT) for HIV.
Isaías Prestes, Suzi Camey
Departamento de Estatística
Universidade Federal do Rio Grande do Sul
Rio Grande do Sul, Porto Alegre
Brasil Av. Bento Gonçalves, 9500