A little Christmas gift for @Nadia , @Noemi and whoever is interested to keep an eye on the shape of the conversation in opencare.
I have made a Tulip perspective that contains no graph data, only Python scripts. If you download it and follow the instructions (which consist mostly of launching the scripts one after the other) you obtain nice-looking visualizations of the opencare social network, based on live data. The first script gets the data through Edgeryders APIs.
It’s on GitHub: https://github.com/opencarecc/rapid-social-networks . Detailed instructions are on the repo’s wiki. As an added feature, a script creates a graph where the community managers (Noemi, Nadia and myself) are removed. This tells you how much peer-to-peer interaction versus person-to-project interaction there is. Results are encouraging: 90% of users are connected to the giant component, even when you disconnect all three community managers.
Yesterday the network looked like this: