NetRunner card recommendation engine

Built on-top of NetRunnerDB user decks, a card recommendation engine.

How to setup from scratch

What this does is import all the cards&decks from NetRunnerDB into PostgreSQL, amoritize the data for neo4j batch-import, and finally imports the data to neo4j.

• git submodule update --init
• bundle install

** You'll need PostgreSQL to continue from here, if you don't you'll need to rewrite the export to neo4j_csv task, as it relies on PostgreSQL COPY command.

• cp config/database.yml.sample config/database.yml Make sure the database.yml is configured properly to your local PostgreSQL.
• rake db:create db:migrate
• rake import:all - This task will import all the cards & decks from NetRunnerDB to your DB.
• rake export:neo4j_csv - This task will export all the cards & decks for Neo4j batch-import.
• rake neo4j:install - Using Neography's task to install neo4j
• rake neo4j:start - You can use rake rake neo4j:start / rake neo4j:stop to toggle on/off the server.
• brew install maven - We need Maven to run the import script.
• ./scripts/import_csv.sh - Using batch-import to import to neo4j the graph.
• rake neo4j:restart - We have to restart the neo4j db after the batch import.

How to update cards.json

• rake generate:card_score - Use this task to update the recommendations score in PostgreSQL (Used in export to json)
• rake export:json - Use this task to build the cards.json file which stores all the cards metadata (including recommendations)
• ruby app.rb - Starts the Sinatra server

At this point if everything ran smoothly, then you can play around with the server at http://localhost:4567