Hi, this is Dinghao! The Network page is now live on The Zone, joining the previously finished Glossary and Reference.

I have used, and will continue to use this page to showcase my analyses of the various paranoia webs in Gravity’s Rainbow. As of now, the section is still WIP, and contains only the interactive (of course) Character Atlas. Some other analyses have been completed and I plan to port them in as interactive plots in the coming week.

Character Atlas

This is a force-directed graph of a curated set of 178 named characters, with edges weighted by co-presence across the novel. Instead of using raw counts for the weighting, the Atlas uses the minimum mention-share in each shared episode, so that a two-character episode where both split attention equally contributes more than a crowded ensemble scene where one character surfaces briefly among dozens. This is a common methodological choice that I continued to use for other analyses which will be visualised in the following week.

Analysis was supported mostly by Python’s NetworkX. Node positions are pre-computed with Fruchterman-Reingold (x.spring_layout()), community detection uses the Clauset-Newman-Moore method (nx.community.greedy_modularity_communities()) yielding five communities that correspond recognisably to narrative arcs: ‘The Zone Squad’, ‘London Counterforce’, ‘The Rocket’, ‘Peenemünde’, and ‘Afterlives’.

Using the interactive GUI, you can filter by Part and hover to inspect individual nodes and their neighbourhood.

Methods

I am developing a separate methods page to document entity recognition, edge construction, community detection, layout, and visual encoding in full. The page is not yet live, but will arrive together with the finalised Network page (hopefully).

Until next time!