|
Activity Number:
|
644
- Statistical Computing on Parallel Architectures
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 2, 2018 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
|
Abstract #326477
|
Presentation
|
|
Title:
|
Automatic Parallelization of R Code
|
|
Author(s):
|
Clark Fitzgerald*
|
|
Companies:
|
University of California, Davis
|
|
Keywords:
|
R;
metaprogramming;
parallel programming;
big data;
databases;
automation
|
|
Abstract:
|
Conventional systems for parallel programming require users to modify existing serial code to take full advantage of a platform's computational capabilities. In this presentation we consider methods to automatically infer the potential for parallelism based on R language idioms and code structure. A system can then automatically transform the serial version of the code into equivalent new versions that offer improved performance. The parallelism could come from simple replacement of functions, or from more elaborate code rewriting, or even from mapping R code into a totally different technology.
|
Authors who are presenting talks have a * after their name.
