SPRINT: A New Parallel Framework for R
Author Information
Author(s): Hill Jon, Hambley Matthew, Forster Thorsten, Mewissen Muriel, Sloan Terence M, Scharinger Florian, Trew Arthur, Ghazal Peter
Primary Institution: EPCC, The University of Edinburgh
Hypothesis
A method of exploiting HPC systems using R, without requiring expertise in parallel programming, is necessary to analyze genomic data effectively.
Conclusion
SPRINT allows biostatisticians to focus on research problems rather than computation, making it easier to utilize HPC systems.
Supporting Evidence
- SPRINT reduces computation time by more than three times when using eight processors compared to one processor.
- The framework allows for easy addition of parallelized functions to R.
- SPRINT is designed to be user-friendly for biostatisticians without requiring knowledge of parallel programming.
Takeaway
SPRINT is a tool that helps scientists use powerful computers to analyze large amounts of data without needing to learn complicated programming.
Methodology
The SPRINT framework is an R wrapper that enables the use of parallelized statistical algorithms on HPC systems.
Limitations
Functions must be re-implemented for SPRINT, which requires significant effort.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website