Graph Visualisation of Airport Connectivity

A visualisation depicting the connectivity of airports around the world. Objective is to get an overview of what airports are more connected, or to what cluster of airports they are most connected to. This will allow one to determine if an airport naturally is better linked with its geographical neighbour airports (on continent), or if an airport is better linked with airports outside the near geographical region.

The data is from 2014 and contains more than 60000 individual flights and 3165 airports. All the airports are loaded as nodes, where the connections defines weight of the edges. The nodes will be sized according to the degree of connectivity, where the edge will be sized according to the count, i.e. number of connections.

First the graph is loaded to which the force directed algorithm is applied. This will cluster together well-connected airports, to which the modularity algorithm is applied that allows the nodes to be coloured according to its sub-network; community structure measure.

Most interesting insight is that London Heathrow Airport is better connected with the continent of North America than with Europe. A number of small airports in Europe are better connected to the greater Russian area than with Europe; Federico Fellini International Airport and Karlovy Vary Airport. Check out the visualisation here.