calculateScenarios: Simulate multiple scenarios for the college admissions...

In tobiasreischmann/matchingmarketsevaluation: Evaluation of the matchingMarkets projekt

Description

This function simulates multiple scenarios for the iterative deferred acceptance mechanism with ties, implemented as stabsim3 within the matchingmarkets package. The results can be used to analyse the number of rounds necessary for the market to be cleared up to a specified threshold.

Usage

calculateScenarios(scenarios, nruns = 10, nworkers = detectCores(),
  seed = NULL, fullresult = FALSE)

Arguments

scenarios	list of lists containing the different scenarios.
nruns	integer indicating the number of markets to be simulated (results are averaged over all simulated markets).
nworkers	integer number of workers generated for the parallel package.
fullresult	boolean if true not only the aggregated rounds of iterations it returned but the full object of each run.

Value

calculateScenarios returns a list of lists, which contains the following fields

occupancyrate	double indicating the ratio of #students/#availableplaces
nStudents	integer indicating the number of students per market
nColleges	integer indicating the number of colleges per market
threshold	double influencing the number of decentrailzed rounds played. The mechanism terminates if the ratio of places, which are different in comparison to the finished mechanism are below this percentage value.
areasize	integer indicating the length of the grid used for the horizontal preferences.
horizontalscenario	integer (0,1,2) indicating which colleges uses horizontal preferences in their ranking (1=>all, 2=>only public colleges, 3=> none).
conf.s.prefs	vector representing the size of the tiers for students' ranking lists
quota	double between 0 and 1 indicating the percentage of private facilities

Author(s)

Tobias Reischmann

Examples

## Simulate a set of different scenarios and return the average number of decentralized rounds played.

elem1 <- list(occupancyrate = .8, quota = .3, nStudents = 2700, nColleges = 600,
              areasize = 7, conf.s.prefs = c(3,7,10,10), horizontalscenario = 1)
elem2 <- list(occupancyrate = .8, quota = .3, nStudents = 600, nColleges = 200,
              areasize = 6, conf.s.prefs = c(2,5,6,7), horizontalscenario = 1)
elements <- list(elem1, elem2)
scenarios <- lapply(elements, function(elem) {
   lapply(c(0.2,0.5), function(x){
      elem$threshold <- x
      elem
   })
})

xdata <- calculateScenarios(scenarios, nruns=2)

tobiasreischmann/matchingmarketsevaluation documentation built on April 25, 2020, 12:58 a.m.