Introduction | Data | Exploratory Analysis | Interactive Visualizations
Faceted Oscillation Propagation Visualization
This visualization is the primary focus of our efforts. It provides an interactive visual representation of the traffic flow oscillations in three facets. The top facet is a heatmap that the displays the flow oscillations overlaid with the mean speed on the axes of time and location. Using this facet, it is possible to identify patterns of oscillation propagation skew as related to changes in speed. Another facet gives a geographic overview of sensor stations and the oscillation magnitude (color channel) at a single time point. Using the animation controls, it is possible to view the wave of oscillation magnitudes move across the stations. The final facet displays the mean oscillations and mean flow for a single station.
As the title implies, the geographic oscillation propagation visualization provides a focus on the positional aspect of the oscillation magnitudes over time. The time controls (animation) allow the user to view waves of the magnitudes move through the highway system. This visualization is a more performant version of the geographic facet in the Faceted Oscillation Propagation Visualization above.
Geographic Oscillation Propagation
Traffic Classification Heatmap
This visualization provides a heatmap representing a heuristic classification of sensor data (limited to I-5 South for all of 2015) on the axes of time and location. The visualization allows exploration via the mean over a range of dates, mean of a single date, or the mean over a range of dates for a particular day of the week.
Station Oscillation Visualizaiton
This visualization presents a multi-faceted way to explore the oscillations of a set of sensor stations (limited to I-15 South). The first facet gives a geographical representation of the sensor station location and its mean flow. The next facet displays the smoothed mean flow of the selected station over the course of an average day. The last facet displays the oscillations extracted from the mean flow.
Oscillation Maxima Visualization
This visualization was created to explore the propagation of the peak oscillation magnitudes for a series of stations (limited to the San Diego area in 2015). Each station is shaded based on the time when the highest magnitude oscillation occurred. The results of this visualization prompted the development of the Oscillation Propagation Visualization described above.
The sensor health visualization below provides an interactive exploratory approach for analyzing missing, outlier, and imputed data within the data set. This is accomplished via a set of multi-faceted dashboards, each focusing on a different aspect of the sensor data.
The "Raw" dashboard provides an interactive display of the raw data (flow, occupancy, and speed) including many options for data selection and filtering. As the "Per Lane Analysis" dashboard suggests, it provides a series of facets for exploring the relationships of speed, flow, and occupancy across individual lanes. The "Obser Health Per Partition" gives a quick overview of the average health of all stations as a function of time. The final dashboard, "Health", provides the ability to visualize the health of stations individually or all stations in a single freeway direction.
| Link | Description |
|---|---|
| WiggleVis | Visualization of Wiggle and Speed |
| SegmentVis | Visualization of Wiggle without heatmaps |
| HeuristicClusterHeatmap | Heatmaps to exploration classification of traffic in Tableau |
| Wiggles_AllStations_Weekdays | Tableau Dashboard to visualize difference in wiggle flow and mean wiggle |
| Traffic_Wiggles | Tableau Dashboard to visualize the average smoothed vector over a freeway |
| Wiggles_by_min | Tableau sensor station max oscillation map |
| SensorHealth | Tableau Dashboard of sensor health |