Scatterplot matrices have always been an effective way to view one distribution versus another. The next version of Dex will also have this capability.
In a scatterplot matrix, numerical values are graphed versus one another. In the following scatterplot, we have plotted a categorical value called “region” versus 3 demographical values associated with each region:
- Age – The age of the resident within that region.
- Income – The income of that resident.
- Children – The number of children of that resident.
There are 4 different regional categories which are treated as series and color coded in the distribution:
- INNER_CITY – Lives within the city.
- TOWN – Lives in a town.
- RURAL – Lives in the country.
- SUBURTAN – Lives on the outskirts of the town.
The image for the scatterplot matrix is below, you can click on the image to get to an interactive version:
Within each scatterplot cell, you can also select the region of interest. Each of the corresponding cells will grey out or ghost the unselected entries. You can then begin to see relationships across the distributions. The selection window can be dragged to dynamically select other regions of interest.