You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am running the Montreal full population matsim scenario, and I have encountered a problem with iteration running time, in starting iteration it was taking at least 2 hours but later it was reduced to 30 minutes do anyone have any idea why in later stages replanning is taking very less time. Is this general behaviour or am I making some error, please let me know.
I set the simulation for 500 iterations with these strategies and corresponding weights:
SubtourModeChoice = 0.05D
ReRoute = 0.1D
ChangeExpBeta = 0.85D
TimeAllocationMutator_ReRoute = 0.05D
The text was updated successfully, but these errors were encountered:
Hello everyone,
I am running the Montreal full population matsim scenario, and I have encountered a problem with iteration running time, in starting iteration it was taking at least 2 hours but later it was reduced to 30 minutes do anyone have any idea why in later stages replanning is taking very less time. Is this general behaviour or am I making some error, please let me know.
I set the simulation for 500 iterations with these strategies and corresponding weights:
SubtourModeChoice = 0.05D
ReRoute = 0.1D
ChangeExpBeta = 0.85D
TimeAllocationMutator_ReRoute = 0.05D
The text was updated successfully, but these errors were encountered: