On this page... (hide)
STARS comes with a built-in example project that we will use to illustrate some of the core functionality. Starting STARS under Windows is done through double-clicking its icon. From there, the example project can be loaded by selecting
The example project has 72 years of data for the 48 lower US states. The key variable we will be analyzing is
To begin our exploration we first we carry out an analysis of the spatial autocorrelation for each year in our sample. This is done by selecting
The view in the upper left if a quinitlie map for the per capita income values in the year 1929. Next to that is the Moran Scatter Plot for the same year. The lower right view is a Box Plot for the incomes in that year and the lower left view is a Time Series plot of the global Moran's statistic for each of the periods in the study.
The four views provide insights on the spatial, temporal and distributional dimensions of the data. To further explore the relationships between these dimensions, STARS implements a number of types of interactive visualization. The first one we illustrate here is referred to as brushing. Brushing allows the user to select observations in an origin view and to see the associated observations in the destination views. To turn linking on the user must first give one of the views focus. This view then becomes the origin view. Next the user issues a
While linking gives the user the ability to locate observations in different views simultaneously, a continuous form of linking, referred to as spatial roaming allows the user to draw a selection window that can then be slid around an origin view. To start the selection rectangle the user presses