[Previous] | [Session 76] | [Next]
E. D. Feigelson, J. Babu (Penn State), A. Mahabal, S. G. Djorgovski, M. Graham, R. Williams (Caltech), R. Nichol, D. Vanden Berk, L. Wasserman (CMU)
A long-standing limitation on the range and sophistication of statistical analysis of astronomical data has been the paucity of non-proprietary software. Our StatCodes Web metasite (http://www.astro.psu.edu/statcodes) has provided links to statistical programs and packages useful to astronomers and other physical scientists, but this collection is heterogeneous and quite incomplete. A more comprehensive and coherent data and statistical analysis environment has recently emerged in the open source R package (http://www.r-project.org) and its associated CRAN archive of add-on packages (http://lib.stat.cmu.edu/R/CRAN). Together, they provide hundreds of statistical functionalities, both simple and advanced, in a programmable data analysis environment with graphics and flexible links to external programs, languages, and databases. We outline these capabilities covering multivariate analysis and classification, parametric and non-parametric tests, regression and smoothing, survival analysis, time series analysis, and spatial analysis. Some of these R functionalities are implemented in our VOStatistics Web service. This work is supported in part by the NSF grant DMS-0101360.
[Previous] | [Session 76] | [Next]
Bulletin of the American Astronomical Society, 36 #2
© YEAR. The American Astronomical Soceity.