ABSTRACT
This chapter is the second of two that outline the programming methods we employed to generate community network graphs. This chapter introduces ways in which attribute data can be attached to the vertex and edge data structures. This can be useful when applying weight to the network graph, so as to show how the community it serves to represent ‘distorts’ its properties, pushing towards certain attractors and pulling away from others. We also propose a way of dealing with uneven community data, reflected in missing attributes, which is common to this kind of research. We propose a method for sharing weights among vertices that are neighbours within network clusters. This means that weighted vertices (that is, vertices that have been described by the participants), might ‘influence’ non-weighted vertices (those that have not been described). We make a convenient assumption that an unweighted vertex that is adjacent to a weighted vertex bears some reflection of the weight value. We offer one example of a community spatial network that has been generated based on this principle for weight-sharing.
