In Reply Refer To:

Mail Stop 440

October 21, 2013

Memorandum

To: GS-W All

From: Nathaniel Booth /signed/
Chief, Office of Water Information

Subject: Transition of Support for Statistics Software From S+ to R

The purpose of this note is to announce the switch in support from S+ to R by the Office of Water Information (OWI). If you use S+ for data analysis, you need to know of these changes. If you use any other statistical package for analysis or graphics, please familiarize yourself with the new resources in R.

This announcement is the culmination of an effort by the R transition team, which has worked for the last two years to provide a relatively smooth transition from S+ to R. The decision to switch from S+ to R was based on cost, ease of transition, and relevance. Briefly, R is freely available, similar to S+, and has a worldwide user base, which will facilitate collaboration with colleagues within and outside of the U.S. Geological Survey (USGS).

It is important that the switch to R be made as soon as possible for all users. The contract for support of S+ will continue through the end of the 2013 calendar year. At that point, support from TIBCO, which owns S+, will stop, but S+ will continue to work on any computer on which it is installed. Users should be able to finish analyses that they have started, but new projects should use R.

The instructions for installing R on PCs are available at: http://bwtst.usgs.gov/apps/R/index.html

Information about R and training and information for developers are available at: /usgs/software/Statistics_Graphics/R/. Two types of training are offered, one is a three-day instructor-led course “Basic Hydrologic Data Analysis Using R,” the other is a self-paced tutorial that focusses on the R language and can be a good follow-up to the instructor-led course or as an introduction to R for those who are more familiar with computer programming languages. Contact Dave Lorenz (lorenz@usgs.gov) for more information or to schedule one of the courses. The instructor-led class can be scheduled in any Water Science Center (WSC) for any time agreed by the instructor and the WSC.

The R Transition Team has identified three general types of users—the casual user, the data analyst, and the developer. The casual user is one who conducts periodic and relatively straightforward statistical analysis. The data analyst conducts more frequent and complex statistical analysis and therefore may have a greater need to learn and use the R language. Finally, the R developer writes new statistical applications or extensively adapts existing applications for new purposes.

We have developed resources for each of the three types of users. For the casual user, a simple GUI has been developed; this GUI will continue to undergo development and documentation as user feedback is provided. The GUI is relatively straightforward and is geared to the level of understanding corresponding to the USGS class “Statistical Methods for Environmental Data Analysis.” The first day of the class “Basic Hydrologic Data Analysis Using R” is also devoted to the GUI.

For the data analyst, a comprehensive suite of functions have been developed to facilitate the import and manipulation of data from the National Water Information System (NWIS) as well as the development of specialized applications for hydrologic data analysis. The second and third days of the class “Basic Hydrologic Data Analysis Using R” are devoted to a basic understanding of the R language. Also, a self-paced tutorial on using R for hydrologic analysis has been started and will be developed for more advanced topics.

Some work has been started to assist developers, including guidelines for producing consistent code from USGS authors and guidance on setting up formal R packages. This work will continue as user feedback is provided and as a network of developers evolves.

Please contact the lead of the R Transition Team, Dave Lorenz (lorenz@usgs.gov), for any questions about R or the transition from S+ to R.