zhaowill / rdea

Data Envelopment Analysis (DEA) package for R with robust unbiased methods.

rDEA

Data Envelopment Analysis (DEA) package for R, estimating robust DEA scores without and with environmental variables and doing returns-to-scale tests.

Installation from CRAN

The package can be installed from the official R CRAN by

install.packages("rDEA")

Under Linux make sure you have GLPK installed before, e.g., in Ubuntu you need

sudo apt-get install libglpk-dev

Or if you need to compile GLPK set following environmental variables before installing rDEA

export CPATH=/path/to/include/glpk
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/path/to/lib/glpk

Installation from git

Using devtools our package can be installed by (make sure you have libglpk-dev installed)

library(devtools)
install_github("jaak-s/rDEA")

In Linux and Mac another option is to use command line (make sure you have slam, truncreg and truncnorm R packages installed)

git clone https://github.com/jaak-s/rDEA.git
R CMD build rDEA/
R CMD INSTALL rDEA_*.tar.gz

DEA Example

Robust DEA (Simar and Wilson, 1998) with input model with included Japan hospital data

library(rDEA)
## loading Japan hospital data
data("hospitals", package="rDEA")

## choosing inputs and outputs for analysis
firms = 1:50
Y = hospitals[firms, c('inpatients', 'outpatients')]
X = hospitals[firms, c('labor', 'capital')]

## Robust DEA with 1000 bootstrap iterations and variable returns-to-scale
di = dea.robust(X=X, Y=Y, model="input", RTS="variable", B=1000)

## robust estimates of technical efficiency for each hospital
di$theta_hat_hat

Testing installed rDEA package

After installing rDEA package you can run included tests by

library(testthat)
test_package("rDEA")