22S:295:002
High Performance Computing in Statistics

Luke Tierney and Kate Cowles

Fall 2007

Information and Resources

• Class meets 9:00-10:00 AM on Thrusdays in Schaeffer 241B.
• Computing resources.
• Class notes:
  • Introduction, August 30, 2007.
  • The Statsitics Cluster, September 6, 2007.
  • Introduction to snow, September 13, 2007.
  • Some snow Examples, September 20, 2007; code examples.
  • More snow Examples, September 27, 2007; kriging example, image processing example, and image data.
  • Using MPI in C programs, October 4, 2007; code examples
  • Brief Introduction to OpenMP, October 11, 2007.
  • Batch Scheduling and Resource Management, October 18, 2007.
  • Managing/analyzing the Netflix data, October 25, 2007; handout; code.
  • A Parallel Approach to Microarray Preprocessing and Analysis, November 1, 2007.
  • The MapReduce Framework, November 8, 2007; mapper and reducer for word counting; mapper for movie rewiew counting.
  • Introduction to PLAPACK, November 15, 2007; some supporting materials.
  • Grid Computing and the TeraGrid, November 29, 2007; some supporting materials.
  • Some changes in snow and R, December 13, 2007.
• Cluster notes:
  • Using snow on the Statsitics Cluster
  • Some simple examples

References

• The Landscape of Parallel Computing Research: A View from Berkeley ( pdf and wiki)
• First chapter of Handbook of Parallel Computing and Statistics, Erricos John Kontoghiorghes (Ed.), 2006.
• PyStream: Stream and GPU computing in Python.
• iPython and parallel computing with iPython.
• Parallel Programming with MPI, by Peter Pacheco, Morgan Kaufmann, 1997.
• PVM: Parallel Virtual Machine by Al Geist et al., MIT Press, 2000.
• Simple Parallel Statistical Computing in R, by A. J. Rossini, L. Tierney, and N. Li, Journal of Computational and Graphical Statistics, 16 (1), 399-420, 2007. Technical report version available from COBRA.
• Parallel Programming in OpenMP by Rohit Chandra et al., Morgan Kaufmann, 2001.
• Patterns for Parallel Programming by Timothy G. Mattdon, Beverly A. Sanders, and Berna L. Massingill, Addison Wesley, 2005.
• Condor home page and manuals.
• MapReduce: Simplified Data Processing on Large Clusters, by Jeffrey Dean and Sanjay Ghemawat, OSDI'04: Sixth Symposium on Operating System Design and Implementation, San Francisco, CA, December, 2004. Available online.
• Wikipedia entry for MapReduce.
• Some MapReduce software: Hadoop from Apache; IBM MapReduce Tools for Eclipse; the Pig project; the Sawzall language.