Ethan Brown
Open-source software developer and statistician
The R Programming Language as a Unified Environment for Data Sonification
Why should auditory display researchers and programmers care about R? R is a free and open source programming language geared towards statistics and data analysis, allowing the analyst to work with data and simulations interactively. In recent 2011 estimates, its academic popularity is only eclipsed by the proprietary statistical software giants SAS, SPSS, and Stata. R has a large and active user base with an exponentially growing number of add-on packages and more blogs and mailing list activity than any other statistical package.
Auditory graphing tools seem like they would fit well into the R community's penchant for experimentation and trying new techniques, yet disappointingly few implementations currently exist. In 2002, Heymann and Hanson sonified the Hybrid Monte Carlo algorithm and developed new tools for working with sound in R, but neither of the authors nor the general R community has followed up on this. This is not surprising, given the lack of convenient tools accessible from within an R session that allow high-level manipulation or mapping of sound. Erich Neuwirth provides an initial attempt of using Csound from R with his Rcsound package. However, this provides just a few basic functions for sonifying data. It has not been updated since 2004 and Internet searches for the package only come up with a single publication and Neuwirth's own mailing-list comments, suggesting little interest.
The experimental playitbyr add-on package for R attempts to fill this gap. playitbyr provides a high-level interface from the R command line to map data and statistical summaries to sonic parameters, with its syntax modeled after a popular graphics add-on package called ggplot2. Realizations are currently provided through simple sine wave, MIDI, and Csound interfaces, but much work remains to be done on providing a usable, portable toolkit.
Questions
- What are the most appropriate sonfication tools to provide in a statistics package?
- Are there problems in data analysis can be best addressed by sonification, and what is the best way to serve those needs?
- Who has successfully been using sonification for statistical analysis and modeling?
References
Neuwirth (2001). R sings - or using R to sonify data. Proceedings of the 2nd International Workshop on Distributed Statistical Computing, Vienna, Austria. |
Heymann and M. Hansen (2002). A new set of sound commands for R; sonification of the HMC algorithm. ASA Proceedings, Statistical Computing Section. |
Wickham (2009). ggplot2: elegant graphics for data analysis. Springer New York. |
Back to ICAD2011 Think Tank
Back to Nick Bearman
Next to Carrie Bruce