Finally, the macro transfer was run again, this time using the functionality to generate a report of the transfer matrix at nodes 423 and 454. The report obtained for node 423 is shown below. The column egress denotes trips which alight at node 423 and leave it through an auxiliary transit link. The row access denotes trips that get to node 423 through an auxiliary transit link and then embark on a transit line. The other rows and columns are labels with the corresponding transit line name and a direction indicator (">" or "<").
Conclusions
We have shown that the computation of consistent line-to-line transfer matrices is theoretically possible, and also practically feasible, with the standard EMME/2 software release.
The macro we developed for this purpose has been written to be as independent of the specific application as possible. In specific contexts, however, it is possible to adapt the same ideas to particular project requirement. Instead of computing transfer matrices on the level of individual transit segments, it would e.g. be possible to compute aggregate transfer matrices by line, link or mode groups. Of course the more aggregate the resulting transfer matrices are, the smaller the required amount of computing resources.
The basic idea of adding this kind of additional structure at transfer points can also be exploited for other purposes than just the computation of detailed transfer information. Since the various types of transfers can now be channeled to different nodes, these c
