Graphics Reference
In-Depth Information
is saved, apply the Label Adjust layout for a second or two to push apart
overlapping labels a bit for better readability.
If you switch to the Preview task mode, you should see a basic view of flight
patternsacrossthecountry.Thesizeofthelabelsindicatesthesizeofairport
by number of domestic flights, and thickness of links indicates the number
of flights between airports. You may need to select the rescaled option on
links and increase their thickness to a higher value (such as 4 or 6) to better
see the results. Turn the arrow heads off by reducing their sizes to zero.
Similarly, remove nodes from the view by setting their opacity to zero.
It's finally time to map flight delay visually to the representation. For this,
you will be using color. Return to the Overview task mode and, using the
Ranking tab, assign average flight delay to the default color scheme. Do this
for both nodes and links. Back in the Preview task mode, select the label
color option, which inherits the original node color, and give the labels a
thick border of five black pixels. Refresh the preview to see the finished
result, as shown in
Figure 13-10
.
Viewing the visualization, you can immediately see best and worst
performers, as well as some overall geographic patterns. Among the worst
are the two main Chicago area airports, O'Hare (ORD) and Midway (MDW),
along with Denver (DEN) and Newark (EWR). A seasoned traveler might
recognize these as airports that experience flight delays with more
regularity. Areas around Chicago and Denver seem to be similarly
susceptible to slowdowns, implying perhaps overcrowded airspace, or
weather-related delays. These are December statistics after all. However,
Newark, seemingly an area anomaly, is not so easily explained.
Interestingly, though, one of the worst-performing East Coast airports in
this data appears to be Trenton-Mercer (TTN), a small airport also located
in New Jersey.
