Graphics Reference
In-Depth Information
Potential Graph Data Sources
The previous scenario is an example of graph data existing buried inside
transactions. Graph data can exist in many different ways in other data.
Finding graph data requires that you identify within that other data both the
nodes and the links between nodes. You will find that some nodes and links
may already exist in other data.
Node and Link Data Sets (for Example, Flight Stats)
On rare occasions, there can be data sets that are already nicely organized
with both node data and link data. One example is a flight database (for
example,
http://transtats.bts.gov
)
.
Are you a frequent traveler? Ever wonder which routes have the most
passengers? Or, have you found a route that has a lot of competition
presumably leading to better fares? Have you ever wondered which routes
have the most flight delays? There is a wealth of flight data on government
websites and through the marketing departments of major airlines that can
be graphed to answer these curious traveler questions.
The flight data sets at
http://transtats.bts.gov
consist of a primary data
set containing each flight (that is, the links), and a supplementary data set
containing each airport (that is, the nodes). In the primary data set, each
flight can be indicated by a single record. The record indicates the city pair
(that is, a link), such as ORD-LGA or LAX-ATL. Note that this particular
data has directed links. ORD-LGA is a flight that starts in Chicago's O'Hare
Airport and ends at LaGuardia Airport in New York City and is different
from LGA-ORD, which is a flight going in the other direction. Both links are
valid.
Flight, Origin, Destination, Distance, Duration
AA100, ORD, LGA, 836km, 1:55
DL364, LAX, ATL, 2384km, n/a
AA102, ORD, LGA, 836km, 1:55
QF32, LHR, SYD, 11711km, 22:18
...
