search

CMAP-REPOS / cmap_csvm

CMAP's commercial services vehicle model source code
1 stars 0 forks source link

Check: clustering of stops/outlier TAZs with too many stops #58

Closed ColSmith76 closed 1 year ago

ColSmith76 commented 1 year ago

Are the any clusters of "other" activity stops in locations with very high land use numbers, e.g., large emploers?

Are locations with high numbers of trips reasonable based on land use in those TAZs. Compare total trip ends with total employment, employment by type, households, and then by activity with total employment, employment by type, households to look for any anomalies. Should be reasonably consistent by employment type (some emp types like locations with construction employment are going to have higher numbers of trips. Since the model is not covering long haul freight, freight facility locations should not pop out but light industry, wholesaling, retail distribution etc probably will show up given local delivery, etc

ColSmith76 commented 1 year ago

Model outputs show a few outlier TAZs, related to particular high employment in land uses that attract intermediate stops. Issue for "other" purpose stops appears to be related to the use of food drink employment in the model specification. Need to check the estimation of that model and consider a logged employment variable so TAZs with a high number of employees don't behave in an extreme way.

ColSmith76 commented 1 year ago

The respecification of the intermediate stop location attraction model seems to have resolved this issue.