MATSim Code

playground cleanup, deleted no longer used classes.

got rid of all calls to Plan.getIteratorAct(). removed Plan.getIteratorAct(), PlanImpl.getIteratorAct(), BasicPlanImpl.getIteratorAct(), BasicPlanImpl.ActIterator. One deprected method less in a core-api...

enable choice set reduction algorithm to write just the choice sets

2 additional test cases for testing QueueSim-behavior with Plans where the last Activity has an end time set.

bugfix: if a person has only one Activity, with an end time set, the person was not correctly removed from the simulation (see r6698, missing call to Simulation.decLiving()), so at the end the simulation thought there is still somebody traveling somewhere although there wasn't. Fixed the original issue (see r6698) differently that solves this problem.

Added a check into "advancePlanElement" if this is not the last plan element. Comment in code:

/*
* My Seattle plans do not run any more without this condition. The problem, presumably, is that some plans have only a
* single activity (e.g. "stay at home all day"). Initially, these activities do not have an end time, and there should be
* no problem. However, the strategy modules (presumably the time allocation mutator) introduces activity end times.
* Ironically, there used to be a condition that activities MUST have an end time, which I disabled. I hope we can afford to
* make MATSim more forgiving here. kai, may/01
*/

A question is if I should perhaps abstain from committing into the core. If so, is there a mechanism to submit patches? If so, do
we have enough resources to handle them?

got rid of all calls to Plan.getIteratorLeg(). removed Plan.getIteratorLeg(), PlanImpl.getIteratorLeg(), BasicPlanImpl.getIteratorLeg(), BasicPlanImpl.LegIterator. One deprected method less in a core-api...

bugfix: replace '&' with '&&' as the logical AND was mend, and not the bit-operation.

got rid of some deprecated method-calls, mostly Plan.getIteratorLeg() and Plan.getIteratorAct()