removeAssignmentPairwiseAtLeastComparisons: Remove assignment pairwise _at least_ comparisons

In kciomek/rorutadis: Robust Ordinal Regression UTADIS

Description

This function removes pairwise at least comparisons. For more information see addPairwiseAtLeastComparisons.

Usage

removeAssignmentPairwiseAtLeastComparisons(problem, ...)

Arguments

problem	Problem from which preference information will be removed
...	Comparisons as three-element vectors and/or two-element vectors. Each argument represents comparison to remove. If c(i, j, k) vector was provided a corresponding comparison will be removed. In case where two-element vector c(i,j) was given a comparison of an alternative a_i with a_j will be removed regardless of value of k. If a specific comparison was not found nothing will happen.

Value

Problem with removed comparisons.

Examples

# 4 alternatives, 2 gain criteria, 3 classes, monotonously increasing
# and general marginal value functions
perf <- matrix(c(5, 2, 1, 7, 0.5, 0.9, 0.4, 0.4), ncol = 2)
problem <- buildProblem(perf, 3, FALSE, c('g', 'g'), c(0, 0))

# add comparisons:
# alternative 2 to class at least as good as class of alternative 1
# alternative 4 to class at least better by 1 class then class
# of alternative 3
problem <- addAssignmentPairwiseAtLeastComparisons(problem,
   c(4, 3, 1), c(2, 1, 0))
# remove comparison between alternative 4 and 3
problem <- removeAssignmentPairwiseAtLeastComparisons(problem, c(4, 3))

kciomek/rorutadis documentation built on May 20, 2019, 8:16 a.m.