ror (version 1.2)

utagms: UTA^GMS MCDA solver

Description

Implements UTA^GMS robust ordinal regression: computes either the necessary- or the possible relation. Assumes ascending preferences, i.e. higher criterion evaluation means higher preferability (=better).

Usage

utagms(performances, strongPrefs=NULL, weakPrefs=NULL, indifPrefs=NULL, necessary=TRUE,
strictVF=FALSE)

Arguments

performances

m x n performance matrix with m alternatives and n criteria.

strongPrefs

k x 2 matrix of strong preferences statements (row indices of alternatives in the performance matrix). Each row r is a preference statements meaning that preferences[k,1] is strongly preferred to preferences[k,2].

weakPrefs

k x 2 matrix of weak preferences statements (row indices of alternatives in the performance matrix). Each row r is a preference statements meaning that preferences[k,1] is weakly preferred to preferences[k,2].

indifPrefs

k x 2 matrix of indifference preferences statements (row indices of alternatives in the performance matrix). Each row r is a preference statements meaning that preferences[k,1] is equally preferred to preferences[k,2].

necessary

Whether to compute the necessary relation (TRUE) or the possible one (FALSE).

strictVF

Whether to use sctrictly increasing (TRUE) or monotonously increasing (FALSE) value functions.

See Also

rorsmaa,ror-package

Examples

library(ror)

## Example with 3 alternatives and 3 criteria
performances <- matrix(c(1.0, 1.0, 1.0, 2.0, 1.0, 1.1, 2.0, 0.5, 3.0), ncol=3, byrow=TRUE)
## a3 > a2 (strongly preferred)
strongPrefs <- matrix(c(3, 2), ncol=2, byrow=TRUE)

## Necessary relation
necrel <- utagms(performances, strongPrefs, necessary=TRUE, strictVF=TRUE)
## Possible relation with strictly increasing value functions
posrel <- utagms(performances, strongPrefs, necessary=FALSE, strictVF=FALSE)

## Sanity check, the necessary relation should be
## T F F
## T T F
## T T T
stopifnot(necrel == matrix(c(TRUE, FALSE, FALSE, TRUE, TRUE, FALSE, TRUE, TRUE, TRUE), ncol=3, byrow=TRUE))

## Sanity check, the possible relation should be
## T T F
## T T F
## T T T
stopifnot(posrel == matrix(c(TRUE, TRUE, FALSE, TRUE, TRUE, FALSE, TRUE, TRUE, TRUE), ncol=3, byrow=TRUE))

## Test with a2 ~ a3
necrel <- utagms(performances, strongPrefs=NULL, indifPrefs=strongPrefs, necessary=TRUE, strictVF=TRUE)
## Sanity check, the necessary relation should represent (a2 ~ a3)
stopifnot(necrel[2,3] == necrel[3,2] && necrel[3,2] == TRUE)

## Not run: 
# 
# ## Plot the relation
# plot(necrel)
# 
# ## a3 > a2 and a2 > a3
# strongPrefs <- matrix(c(3, 2, 2, 3), ncol=2, byrow=TRUE)
# 
# ## Error as the model is infeasible
# necrel <- utagms(performances, strongPrefs, necessary=TRUE, strictVF=TRUE)
# ## End(Not run)