…PASs when not used, updating bush tun flows during loading update)

- added test with consecutive PASs single od (working)
- added test with multiple destinations (not working yet, continue with this)

… the wrong way around. This is fixed

- bug fix sLTM: when between iterations a node switches from potentially blocking to non-blocking, its costs must be updated. This was not the case due to its now non-blocking nature. This is now fixed.
- bug fix: sLTM: outflows are now updated on all tracked node incoming link segments, this was not the case.
- added multi-destination test with assertions that fail due to incorrect tracking of bush splitting rates under multiple destinations. To be fixed
- added warning to steady state travel time cost in case outflow is zero and inflow is positive

…not potentially blocking, the node model was still run. This has been fixed

- enhancement sLTM: storing of turn sending flows now by multikey rather than manual combined hash. This is safer

… bush

…litting rates per bush. Continue with flow shift implementation based on identified labelling, then continue with loading update which should track sending flows per label to correctly determine splitting rates per label as well

…h based added. NOT TESTED. We track flows per bush per label where labels are created such that at merges flow are tracked separately such that when diverging the flows are split by label instead of across all flow. This means that flow originating from a different location before the merge can be impeded differently and the diverge would be able to recognise this as flow is stored separately. Without this labelling this would not be possible. It effectively is a smarter way of storing destination based information in the bush

- removed manual combinedhashkeys as map key with multikeymaps from apache commons instead

…est now succeeding.

…compute proportions for the PAS diverge differently. Before these proportions were not yet calculated at all. Now this is added, but not yet tested.

…t code using vars as well. Odtest1 now working again. Continue with SingleOdtest2.

…eod2 test is just rounding? Check and fix then move on with multi-od

…o develop

…ssignment.java)

- updated odtest2 for sLTM. Now passing.

…c final member

…an immutable value)

…nation testcase now working

- added test for multi-od bush-based sLTM
- multi-(o)d bush based tests now succeeding and assertions added

…ion to be more generic (not done)

- shortest bush to be used in sLTM bush-based
- added grid network generator
- added grid based bush sLTM test (not done)

…t path. Not working yet

- support for destination labelled bushes. This is the simple labelling mechanism that in due time is to be replaced by more intelligent labelling
- destination labelled bushes tests for single or and multi destination work and succeed, multi-origin and grid-based tests need to be developed further
- smart labelling prototype is partly implemented in the form of smartLabelling flow shift executor. To be expanded on in the future (created separate issue for this)

…e to include apache commons dependency there

…nversions of service networks

- adding support for virtual mapper in planit component id mappers

… identification of stop locations rather than link based. Improved mode checking and improved location checking considering the extreme nodes knowing the preferred location of access linksegments relative to their access node (whereas with links this is not known and therefore more error prone)

- minor stability improvements

…n contents