Skip to content

Latest commit

 

History

History
 
 

meetups

Contents

Meetup Analysis

file notes
extract.py script that combines the various .json files obtained from the meetup crawlers and outputs to 4 .json files in /output. Warning: json does not support integers as dictonnary keys, so there are type inconsistencies (see here)
extract_pickle.py same as extract.py except the outputs are pickle files, bypassing the json problem.
find_new_members.py script computing 3 metrics and outputting lists for new members in /output
graph_walk.py script producing the metrics from find_new_members + 4 other ones. Outputs to output/metrics.csv. Uses the Sliced_graph class in sliced_graph.py and requires previous use of extract_pickle.py

Visualisation

see also the content of the html/ subfolder

file content produced by
common_topic_graph.json groups linked by common topics topics.py
force.json idem meetup_graphs.py
common_member_graph.json groups linked by common members topics.py
member_force.json idem meetup_graphs.py
group.json nodes = BruDataSci + members + similar groups (common members), edges +-= membership meetup_graphs.py
topics_and_groups_graph.json nodes = topics + groups, edge = 'topic of' topics.py
topics_graph.json topics linked by common groups topics.py

foo_tech.json = same as foo.json restricted to tech meetup groups, produced by topics_tech.py

N.B.: "A linked by common B" means that an edge between nodes A1 and A2 is drawn whenever the overlap between A1 and A2 is substantial (substantial can have different meanings / thresholds for different graphs)

Technical N.B.: in foo.json (non-tech) there is only a nameattribute which is displayed on mouseover only. in foo_tech.json, the name attribute of a node is a shortened version of the fullname and is displayed on top of every node in index_tech.html.