Doing Interesting Things With Your Rails Data

Transcript

1. Doing Interesting Things With Your Data SFRails August 1st, 2013
2. WHO AM I? • Serena Wales • @gangleton on Github and Twitter • 2012 Code for America fellow • CTO of Textizen • Rails dev
3. We are surrounded by data
4. Logs And Exhaust Data
5. Other people’s data!
6. What can you do with the data you’re already collecting?
7. And how do you make it useful for the people using your application?
8. Medium Data
9. Why use ?
10. 3 Approaches
11. • Front End • Back End • Way Back End
12. Front End
13. Explore your data • Gauss • https://github.com/wayoutmind/gauss • All kinds of fun statistical functions • Underscore.js Collection and Array Methods • http://documentcloud.github.io/underscore/ • Like Enumerable, but different
14. Explore your data • http://misoproject.com/dataset/ • Handles loading, parsing, sorting, and querying • Great if your underlying data is changing regularly
15. And make it visually useful • Flot • http://www.flotcharts.org/
16. And make it visually useful • D3 • http://d3js.org/
17. Front end: Pros • Great visualization options • Use tools developed for rich user experiences • Works with dynamic data • Can build in interaction and allow people to customize their views
18. But not always great • Too much processing or data can lock your interface • If you’re not using most of the data, that can be a lot of extra traffic and memory
19. Back End
20. ActiveRecord does things! • ActiveRecord::Calculations • http://api.rubyonrails.org/classes/ActiveRecord/ Calculations.html • Average, Maximum, Minimum, Sum
21. ActiveRecord does more things! • ActiveRecord::QueryMethods • http://api.rubyonrails.org/files/activerecord/lib/ active_record/relation/query_methods_rb.html • Group, Having, Where
22. For more robust analysis... • descriptive_statistics • statistics2 • SciRuby • Statsample • NMatrix
23. Make charts in Ruby • Rubyvis • https://github.com/clbustos/rubyvis
24. Make charts in Ruby • Googlecharts • https://github.com/mattetti/googlecharts require  ‘googlecharts’ data  =  [1,  1.2,  1.7,  1.5,  0.7,  0.3] Gchart.bar(:data  =>  data)
25. Make charts in Ruby • Gruff • https://github.com/topfunky/gruff require  ‘gruff’ data  =  [1,  1.2,  1.7,  1.5,  0.7,  0.3] g  =  Gruff::Bar.new g.data  :heights,  data g.write(‘chart.png’)
26. Where do I put this stuff? • Not in your view. • Probably not in your model • Use more classes
27. Load time > 3 minutes • Optimize for reading • Cache results • Recalculate after changes
28. Back End: Pros • Object-oriented organization • Access whatever you need • Take advantage of your data store
29. Back End: Cons • Graphing • Underdeveloped or underfeatured libraries • Page rendering performance
30. Way Back End
31. For when your data is really, really complicated or you have a lot of it
32. Doing complicated things with numbers? • Scientists seem to like R • RSRuby • https://github.com/alexgutteridge/rsruby • Rook • http://cran.r-project.org/web/packages/Rook/ index.html
33. Have lots of numbers?
35. The Future!
36. Thanks for listening! @gangleton www.textizen.com