Analytics to the masses by JOSE LUIS LÓPEZ at Big Data Spain 2014

In this talk, we will discuss our approach to bring large scale deep analytics to the masses. R is an extremely popular numerical computer environment, but scientific data processing frequently hits its memory limits. On the other hand, system to execute data intensive tasks like Hadoop or Stratosphere are not popular among R users because writing programs using these paradigms is cumbersome. We present an innovative approach to overcome these limitations using the Stratosphere/Apache Flink big data platform by means of a R package and ready-to-use distributed algorithm.

Transcript

1. BIG DATA ANALYTICS TO THE MASSES JOSE LUIS LÓPEZ PINO DATA ENGINEER GETYOURGUIDE
2. Big Data Analytics to the masses Why it has failed and how we can fix it Jose Luis Lopez Pino
3. Who am I? BI Consultant Large-Scale & Distributed Founding Data Engineer
4. Big Data is like Tourism But if you aren’t an expert, you can’t make the most of it It seems easy to do
5. Struggle to analyze Big Data Harlan Harris, Sean Murphy, and Marck Vaisman. Analyzing the Analyzers: An Introspective Survey of Data Scientists and Their Work. O’Reilly Media, Inc., 2013 Also: Sean Kandel, Andreas Paepcke, Joseph M Hellerstein, and Jeffrey Heer. Enterprise data analysis and visualization: An interview study. Visualization and Computer Graphics, IEEE Transactions
6. Tools Volker Markl. Breaking the chains: On declarative data analysis and data independence in the big data era. Proceedings of the VLDB Endowment, 7(13), 2014
7. Tools (October 2014) Original: Volker Markl. Breaking the chains: On declarative data analysis and data independence in the big data era. Proceedings of the VLDB Endowment, 7(13), 2014
8. Deep analytics
9. Libraries! We need libraries... Query languages Write your own MR/RDD/Transformations
10. … comprehensive ones!
11. Say it with memes! When you do Deep analytics in small data using R and CRAN packages When you do deep analytics in BIG data using R and CRAN packages
12. When you try to program it using MapReduce When you try to program it using Apache Spark / Apache Flink When you try to use a library scalable to large data sets
13. Can’t we do it better? - Make it similar to normal R programs. - Hide complexity. - Make file manipulation easier. - Part of the computing in the cluster and part of the computer in the client.
14. Our approach
15. Our approach
16. Behind the scenes: Before
17. Behind the scenes: After
18. Without writing significantly different code
19. Competitive or even faster than R native code in small data
20. And it scales
21. Some relevant findings - Transmission time was not significant. - Stratosphere/Flink was competitive in highly iterative programs. - We were not able to do it keeping the code 100% the same. - Ensemble scenarios are the most exciting ones.
22. 4 Takeaways from this talk - We still need to bring Big Data to the right people in the right place. - We need comprehensive libraries. - We need to move data back and forth. - Use a syntax that the users are familiar with.
23. That’s all! - Have you found this talk interesting? - Follow me: @jllopezpino - Interested in a job as SEM Data Analyst (Berlin)? - Ask me for the details: - Are you interested in Data + Energy? - Keep in touch:
24. 17TH ~ 18th NOV 2014 MADRID (SPAIN)