Benchmarking do DEA para o Juarez

Benchmarking_contribution_of_each_DMU.R

# Install the Benchmarking package if you haven't already
# install.packages("Benchmarking")

# Load the library
library(Benchmarking)

# Sample data: inputs (x) and outputs (y) for a set of DMUs
# Example from the package documentation
x <- matrix(c(100, 200, 300, 500, 100, 200, 600), ncol = 1)
y <- matrix(c(75, 100, 300, 400, 25, 50, 400), ncol = 1)

# Run the DEA model (e.g., input-oriented, VRS or CRS)
# The 'dea' function returns an object of class 'Farrell'
# We'll use VRS (Variable Returns to Scale) as an example
e <- Benchmarking::dea(x, y, RTS = "VRS", ORIENTATION = "in")
e$eff
# Get the peers for each DMU
# The 'peers' function returns a matrix where each row corresponds to a DMU, 
# and the columns list the indices (or names, if specified) of its peers.
# Efficient DMUs are their own peers.
peers_list <- peers(e, NAMES = TRUE)

# Print the peers
print(peers_list)

# To get the contribution (lambda values) of each peer, you can use the 'get.peers.lambda' function
lambda_list <- get.peers.lambda(e)

# Print the lambda values
print(lambda_list)

The peers(e) function will return the indices (or names) of the efficient DMUs that form the benchmark for each evaluated DMU. 

The get.peers.lambda(e) function provides a list where each element corresponds to a DMU, listing its peers and the associated (\lambda ) weight. The (\lambda ) values represent the intensity variables indicating the contribution of each peer to the inefficient DMU's performance target. 

This approach allows you to identify exactly which efficient units a specific inefficient DMU should emulate to become efficient. 

install.packages("farrell")

https://cran.r-project.org/web/packages/farrell/vignettes/Intro.html

library(farrell)
farrell()