|
Activity Number:
|
304
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #300631 |
|
Title:
|
BigmemoRy: Massive Data, Shared Memory, and Parallel Computing in R
|
|
Author(s):
|
John W. Emerson*+ and Michael Kane
|
|
Companies:
|
Yale University and Yale University
|
|
Address:
|
438 Humphrey St, New Haven, CT, 06511,
|
|
Keywords:
|
R ; memory ; parallel ; C ; data ; data mining
|
|
Abstract:
|
The new bigmemoRy package (described earlier by Michael Kane) allows users to store massive matrices in shared memory in UNIX environments. This has several exciting benefits. First, multiple R sessions on the same machine can share access to the same massive data set, avoiding the need for each session to manage its own copy. Second, it opens the door for development and implementation of parallel algorithms with massive data sets. This package is designed for a wide range of users, ranging from serious C and Fortran programmers to casual R users with no experience with other programming languages.
|
