Embedded Files
Goal: Time series graph that displays the number of human exposure calls by condition over time. The aberration detection algorithm used to detect call volumes >=4 standard deviation units from the mean call volume is a modified version of the EARS C2 statistic known as W2. Expected call volumes per condition are calculated using a 28-day moving average where the two days immediately preceding the index day are skipped.
Actors: Public Health Investigator
Preconditions: A user has been authenticated. A state/zip3 or zip5 has been selected. A date range has been selected and stored in session.
Success condition: A graph is displayed.
Main Success Scenario:
- A graph is displayed with each selected day as a data item with time as the x-axis and count as the y-axis. A line connects all data points.
- An additional line is displayed for the rolling 28-day average for each data point.
- A star is displayed on the data point if the count exceeds 4 standard deviations for that data point.
- A link is displayed under the chart to the selected geographic area. When clicked, the graph popup closes and the map is displayed for the current region with all of the sub regions displayed. For example, the link for the Colorado chart will lead to a map of Colorado displayed with all of the zip3s shaded based on call volume by selected condition.
Alternate Flow #1:
(Step 1): The graph supports zooming in to specific sections so that the data points can be viewed in greater detail. This is built-in behavior to the flot toolkit that is being used to build the graphs.
Notes:
References for calculating standard deviation: hubpages, wikipedia
References for W2 statistic details are found in attached BioSense hospital data user guide.
Related Use Cases:
Mockup:
Page updated
Google Sites
Report abuse