swimr: Visualizing Complex Longitudinal Indicators for a Statewide Integrated Land Use and Transport Model in Oregon
Corresponding Author: Gregory Macfarlane, Transport Foundry
Presented By: Gregory Macfarlane, Transport Foundry
Abstract
The Oregon Department of Transportation operates a statewide integrated model (SWIM) with an exchange-based economic and land use model (PECAS) and an activity-based personal transport model. In order to quickly and easily examine longitudinal trends in economic and transport metrics -- and thereby validate the model's reasonableness and sensitivity -- we develop an R package swimr to generate plots and interactive maps. The package functions query data from one or more model scenarios and output high-level graphical summaries of the model's complex outputs including:
- population and employment
- VMT by facility type
- trips by mode
- trip length frequency distributions
- new floorspace constructed by type
- floorspace occupied by type
- rents by floorspace type
- economic output by sector
These plots are available to view indexed or absolute values from a single scenario or multiple scenarios, or relative difference in output between two comparison scenarios. Function arguments allow the user to specify the geographic aggregation and scope. Examples of some function outputs are attached to this abstract.
Additionally, the package includes RMarkdown templates that can be used to quickly create a comprehensive diagnostic report of a scenario. This report can subsequently be hosted on a website such as at http://52.206.145.100:3838/swim/Reference.html.
The plot functions in swimr are designed to give an analyst quick access to high-level summaries of an immense and complex database with only a few commands. Their delivery within an R package (available on GitHub) makes long-term maintenance, extension, and adaptation feasible.