adea_stepwise: Select an optimal subset of variables for DEA analysis

In adea: Alternate DEA Package

Select an optimal subset of variables for DEA analysis

Description

Stepwise procedure for variable selection in DEA models. This is a back end function for adea_hierarchical and adea_parametric functions. So, it is not for end user use.

Usage

adea_stepwise(
  input,
  output,
  orientation = c("input", "output"),
  load.orientation = c("inoutput", "input", "output"),
  name = "",
  direction = c("backward", "backward/input", "backward/output"),
  load.critical = 0.5,
  max.steps = ncol(input) + ncol(output) - 2,
  verbose = 0
)

Arguments

input	A matrix or a data frame with the inputs of units to be evaluated, one row for each DMU and one column for each input.
output	A matrix or a data frame with the outputs of units to be evaluated, one row for each DMU and one column for each output.
orientation	Use "input" for input orientation or use "output" for output orientation in DEA model.
load.orientation	It allows the selection of variables to be included in load analysis. Its default value is "inoutput" which means that all input and all output variables will be included. Use "input" or "output" to include only input or output variables in load analysis.
name	An optional descriptive name for the model. It will be shown in print and summary results.
direction	The direction in which the variables go in and out of the model. Until now, only backward option is implemented.
load.critical	Minimum values for load.ratios to consider that a variable should be considered in the model. It can be also a vector with two values, the first value input loads and the second for output loads.
max.steps	The maximum number of steps allowed.
verbose	Use 0 for minimal output, only final model. 1 or more to get detailed information of each step. This option affects only to printed output but not the result.

Value

The function returns a DEA model with optimised set of variables.

adea documentation built on March 18, 2022, 7:24 p.m.