The list of available parameters and valid parameter values may vary from release to release. For a list of all available settings available with the version you are working with, run "DefaultConfig.java" in the Metacity-MATSim module.
Configuration File
The configuration file consists of modules. Each module represents a logical parameter group. Numerous MATSim modules can be added to MATSim and configured by specifying the respective configuration file section.
Modules
List of modules to configure in our Prague scenario
global
randomSeed
coordinateSystem
numberOfThreads - for parallel event handling
insistingOnDeprecatedConfigVersion
controler
outputDirectory
firstIteration
lastIteration
compressionType
dumpDataAtEnd
routingAlgorithmType - the Fast versions takes up more memory
overwriteFiles
createGraphs
writeEventsInterval
writePlansInterval
writeSnapshotsInterval
writeTripsInterval
qsim
startTime
endTime - should be specified so the simulation does not run forever for agents with faulty plans
flowCapacityFactor - scales the capacities of the routes accordingly
storageCapacityFactor - tested for flowCapacityFactor^(0.75)
snapshotperiod
mainMode - modes that use routing in qsim (usually car)
linkDynamics
trafficDynamics
usePersonIdForMissingVehicleId
numberOfThreads - for parallel QSim
vehiclesSource - specifies whether only vehicle types are used or more detailed type for each vehicle ID (see Vehicles)
insertingWaitingVehiclesBeforeDrivingVehicles
network
inputNetworkFile - path to MATSim network
plans
inputPlansFile - path to plans/population file
transit
useTransit - specify whether PT should be simulated
transitScheduleFile - path to transit schedule file
vehiclesFile - path to vehicles file
routingAlgorithmType - the type of transit routing algorithm used (default: SwissRailRaptor)
transitModes - for public transit equals to "pt"
facilities
todo
households
todo
vehicles
vehiclesFile - file specifying vehicle types (or more detailed their id)
planCalcScore
writeExperiencedPlans
fractionOfIterationsToStartScoreMSA - 1 or 0, Forcing Scores to Converge
learningRate
BrainExpBeta
lateArrival
earlyDeparture
performing
marginalUtilityOfMoney
utilityOfLineSwitch
waiting
waitingPt
parameterset type="modeParams" - specifies parameters for each leg mode (e.g. car)
mode
constant
dailyUtilityConstant
marginalUtilityOfDistance_util_m
marginalUtilityOfTraveling_util_hr
monetaryDistanceRate
dailyMonetaryConstant
parameterset type= "activityParams" - must be described for every activity in plans/population file (e.g. home, work, shop, leisure, ...). There are also activities specified by the simulation (e.g. car interaction), do not have to be defined explicitly.
activityType
priority
typicalDuration
openingTime
latestStartTime
earliestEndTime
closingTime
planscalcroute
networkModes - leg modes that will be reated by the network router (usually cars)
parameterset type="teleportedModeParameters" - leg modes using teleport (walk, ...)
beelineDistanceFactor
mode
teleportedModeSpeed - uses beeline distance
teleportedModeFreespeedFactor - uses distance from the route on network (not simulated!)
travelTimeCalculator
analyzedModes - usually car
separateModes
TimeAllocationMutator
mutationAffectsDuration
mutationRange - defines how many seconds a time mutation can maximally shift a time
swissRailRaptor
useCapacityConstraints
strategy
maxAgentPlanMemorySize
fractionOfIterationsToDisableInnovation - disables innovation stage after a % number of iterations
parameterset type="strategysettings" - various strategies for running the simulation
strategyName - values [BestScore, ChangeExpBeta, ReRoute, SubtourModeChoice, TimeAllocationMutator_ReRoute]. *Some of these need their own module to further specify the usage.
weight
*module: subtourModeChoice
chainBasedModes - modes that preserve mass
modes
considerCarAvailability
