Chapter 17 Visualising

Now the network has been aggregated, there are decisions to make about:

  • encoding important facts about the network e.g. in terms of colour, width, visibility, etc
  • decoration - other things which make the network easier to read, such as orientation, gaps between variables, etc.

Here we will deal only with the first.

17.1 Conditional formatting

In the Display tab, the user can set almost any visual feature of the network to reflect almost any attribute of the variables, arrows etc.

For example:

  • Hide all variables with some attribute equal to, or less than a certain number, or containing a certain string ….
  • Make the width or transparency of the arrows reflect the number of mentions
  • Colour arrows according to the average age of people mentioning them

etc.

This can seem a bit overpowering so the trick will be to have plenty of sensible pre-sets and templates.

17.2 Hard-coded formatting

(Very debatable:) e.g. 

  • make an arrow dotted if it is coded as a definitional/conceptual rather than causal link
  • put a symbol on an arrow if its strength is less than 0

17.3 Constructing labels and tooltips

Arrow and variable labels and tooltips can be constructed by the user to show e.g.

  • the actual label originally provided by the user e.g. “Income due to pig farming”
  • any attributes available to the app, e.g. list of sources, average age, etc.

So if the user specifies e.g. label, frequency in the setting variablelabel, the variable labels will look like this:

Income due to pig farming
Frequency: 46

17.4 Focus

If diagramfocus is TRUE: This is mainly for interactive viewing, not for printing / exporting. As the user clicks through the different statements in the corpus, the arrows coded on the basis of this statement are highlighted and the others fade. (There is also a “play” button which plays through all the statements like a video.)