Visualise the social graph behind Lanyrd events with an interactive JS graph
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
SocialGraphVis
README
get_friends.py
get_lanyrd_data.py
make_graph.py

README

Visualise a Lanyrd.com events page with a force directed graph (via ProtoVis) as a javascripted web page - shows you an interactive network of who-knows-who at an event. See this write-up for live demos:
http://ianozsvald.com/2010/10/05/visualising-lanyrds-social-connectivity-graph/

python get_lanyrd_data.py
goes to http://lanyrd.com/2010/barcamp-brighton/ and uses regexps to extract list of attendees as twitter screen names, pickles result as lanyrd_attendees.pickle
NOTE you might have to run this several times, it'll use Twitter's 150 API requests per hour quite quickly and for any larger event (e.g. >20 people) you'll run out of the hour's requests. Just re-run it and it'll pick-up where it left off.

python get_friends.py
loads lanyrd_attendees.pickle, it will build attendee_friends.pickle as it goes. 
it calls twitter for each screen_name in the previous step and gets their friends (along with a lot of other json data from twitter). attendee_friends.pickle will be 40mb for Bar Camp Brighton 5. if it runs out of twitter api calls (150/hr) then it'll die without damaging anything, just run it again later (it'll tell you how many mins are left until the api limit is reset) and it'll pick up until complete

python make_graph.py
reads the large attendee_friends.pickle and lanyrd_attendees.pickle, builds links and nodes which is output as JSON to graph.json

cp graph.js SocialGraphVis
copy the generated graph.js file into the HTML directory, then open lanyard.html in your browser to see the graph
customise logo.png and lanyard.html to fit the event you're visualising (it is set to BarCampBrighton5)

Credit:
Idea - Kyran Dale and Ian Ozsvald
Data grab & graph - Ian Ozsvald
ProtoVis/JS visualisation - Kyran Dale
http://IanOzsvald.com
http://ShowMeDo.com/about (Kyran)

Uses:
http://Lanyrd.com
http://vis.stanford.edu/protovis/
Python 2.5 or above