Essentials of High Performance and Parallel Statistical Computing with R (ADDED FEE)

• Activity Number: 214089 - Essentials of High Performance and Parallel Statistical Computing with R (ADDED FEE)
• Type: Professional Development
• Date/Time: Tuesday, August 1, 2017 : 8:30 AM to 5:00 PM
• Sponsor: ASA
• Author(s): George Ostrouchov* and Wei-Chen Chen*
• Companies: Oak Ridge National Laboratory and FDA/CDRH

Abstract

This is an introductory course in high performance and parallel statistical computing, which is essential for statistical modeling when dealing with big data. We introduce fundamentals of parallel statistical computing including the use of the pbdR package ecosystem on larger platforms. We present a broad overview of parallel programming paradigms and relate parallel approaches within R for statistical computation. Practical examples beginning with strategies for speeding up serial R code and continuing with parallel approaches of increasing complexity are discussed. Computing platforms ranging from multicore laptops to medium and even large distributed systems are covered. We bring a coherent approach that is based on established advanced parallel computing concepts from the high performance computing (HPC) community, all within the comfort of R. Basic knowledge of R and statistical computing are assumed.