@JoeJimFlood, @wusun2
For the MVP, we think it's best to have the RSM zones be aggregations of TAZs. Mechanically, this means we will turn up the weight on the "same TAZ" parameter in the zone aggregation method. Making this assumption will allow us to move faster on (a) creating RSM zone roadway centroids and (b) reducing the non-passenger demand matrices.
When we test the MVP, we can assess whether this simplification is something we want to retain in the next version of the software, or relax to allow TAZ splits.
@DavidOry , this is ok with me for now. There could be cases where a customized geography (such as MoHub geography) need TAZ splits. We can re-evaluate this after MVP is completed.
@JoeJimFlood, @wusun2 For the MVP, we think it's best to have the RSM zones be aggregations of TAZs. Mechanically, this means we will turn up the weight on the "same TAZ" parameter in the zone aggregation method. Making this assumption will allow us to move faster on (a) creating RSM zone roadway centroids and (b) reducing the non-passenger demand matrices.
When we test the MVP, we can assess whether this simplification is something we want to retain in the next version of the software, or relax to allow TAZ splits.
Any thoughts at this stage?
fyi @jpn-- , @elias-sanz