Our Daily Graphs written in Ruby and Neo4j

Talk given at RubyConf AR 2014 - Buenos AIres

Transcript

1. Our Daily Graphs
   Written in Ruby and Neo4j
   

   View Slide

2. Hola! / Oi! / Hi!
   • Hanneli - from Brazil
   • CodeMiner 42
   • Coffee (coffee, coffee, coffee, coffee, coffee,
   coffee), Pokémon, Lego and animals <3
   HIRING
   

   View Slide

3. We *WON’T* see here
   • Advanced graph theory
   • Neo4j complete guide
   • Java stuff (Neo4j is written in Java)
   • Social Networks classics example (eeeww)
   

   View Slide

4. Why graphs?
   In a Ruby Conference!
   

   View Slide

5. Example
   How does Codeminer build dev teams?
   

   View Slide

6. Example
   How does Codeminer build dev teams?
   MATCH
   

   View Slide

7. How many times
   have you made
   drawings like these?
   

   View Slide

8. Some history
   • Euler
   • Statistics
   • Patterns/relationship
   • Dependencies
   • Routes
   

   View Slide

9. Persistence
   

   View Slide

10. Persistence
    

    View Slide

11. Persistence
    One of the first CS subjects is…
    

    View Slide

12. Do you remember?
    Queue
    Stack
    Hash
    Tree
    Graphs
    

    View Slide

13. Persistence
    What’s the first storing system we study!
    into databases class?
    a. Queue
    b. Stack
    c. Hash
    d. Tree
    e. None
    

    View Slide

14. Persistence
    

    View Slide

15. Have you ever thought about it?
    

    View Slide

16. When did I notice that?
    

    View Slide

17. Hydropower modelling
    

    View Slide

18. Hydropower in Brazil
    

    View Slide

19. Wanna represent hydropower map
    into relational model
    HYDROPOWER_PLANT
    locale:float
    capacity:float
    name:float
    providing_city_id:CITY
    neighbor_id:HYDROPOWER_PLANT
    population:int
    

    View Slide

20. FUUUUUUU….
    

    View Slide

21. FUUUUUUU….
    • According to population, how should I create
    hydropower plants neighbors?
    • If one plant shuts off, which one should replace it
    without causing damage to the entire system?
    • How should I avoid underuse?
    • Monitoring?
    

    View Slide

22. Double Fail….
    

    View Slide

23. ‘Unleraned?'
    

    View Slide

24. Look at this
    

    View Slide

25. Similar?
    

    View Slide

26. Similar?
    

    View Slide

27. Graphs oriented
    databases
    Because not all data need to live in tables
    

    View Slide

28. Neo4j
    

    View Slide

29. Our plants into graphs
    Itaipu!
    is_neighbor
    Itá
    latitude:33.43
    longitude:32:33
    capacity:1400
    latitude:43.43
    longitude:56:33
    capacity:700
    line_voltage:230
    

    View Slide

30. Our plants into graphs
    Itaipu!
    is_neighbor
    Itá
    latitude:33.43
    longitude:32:33
    capacity:1400
    latitude:43.43
    longitude:56:33
    capacity:700
    line_voltage:230
    Node
    Relationship
    Attributes
    index
    

    View Slide

31. Visual representation
    Itaipu!
    Itá
    

    View Slide

32. Our plants into graphs
    Itaipu!
    feeds
    Foz do
    iguaçu
    latitude:33.43
    longitude:32:33
    capacity:1400
    population:265000
    feeds
    Curitiba
    population:1700000
    

    View Slide

33. How many people does Itaipu feed?
    

    View Slide

34. Traversal - Search
    Itaipu!
    feeds
    Foz do
    iguaçu
    latitude:33.43
    longitude:32:33
    capacity:1400
    population:265000
    feeds
    Curitiba
    population:1700000
    

    View Slide

35. Cypher
    

    View Slide

36. Neo4j with Ruby <3
    neo4j.rb
    Active Record
    

    View Slide

37. Neo4j with Ruby <3
    REST API Wrapper
    Neography
    

    View Slide

38. Neo4j - When should I not use?
    

    View Slide

39. Neo4j - When should I not use?
    • NoSQL Hater?
    • Searches - BE CAREFUL (depth x width)
    • Cloud
    • Node Monitor
    • Infra (backup)
    • Limted RAM
    

    View Slide

40. Obrigada!
    @hannelita
    hannelita@gmail.com
    Questions?
    

    View Slide