R for Monitoring Natural Resources
These pages seek to meet the needs of the National Park Service Inventory & Monitoring for specific guidance on using R for the reporting, analysis, and synthesis of Inventory and Monitoring data. Despite that narrow targetting, most of the material should be useful to other natural resource scientists and managers: especially those with general database skills and knowledge about their datasets. The pages on this site are meant to enhance visitors' learning experience through interactive tools to aid the R language parsing.
The initial impetus for these pages is as a companion to the "Learn R" webinar course led by Paul Geissler of USGS. While the Monday and Wednesday presentations will be from the textbook with textbook (clean & simple) examples, optional Tuesday sessions will attempt to apply those topics to real NPS data and questions. I hope to produce a resource that will be useful for I&M and other natural resource colleagues well after this course ends.
The content has to be strategic, not comprehensive. I'm generating these on the fly from the Learn R course and from my own work. My official workplan does not include these pages, so I can only justify topics that require only incremental work but at the same time will greatly reduce the number of calls for assistance. As any of you complete tasks or projects using R that would be useful to other I&M networks, I would welcome their addition to these pages. Eventually, these need to become community pages for users of R within NPS I&M and the broader natural resource community.
If you look at the navigation bar on the left of this web page, you will see that there are 2 sets of R pages. One set named "Learn R" includes the index page for this course, plus web pages for each session. The other set ("R by Topic") is divided up by topics, for use as references or examples. The topics pages tend to have a bit more content than is included in the corresponding course page, and has content such as graphics centralized as opposed to spread out on several session pages. I have not yet ported most of my content to these pages, so the R by Topic tree gives an indication of what topics will be included there, but most are not yet links.
What you'll find on the pages
Installation & Configuration-- Hints about installation and configuration for users without full administrative rights to their own computers, and for general efficiency.
Data Import & Export-- How to grab data from relational databases, and how to import climate and other netCDF data
Graphics-- Over time this section will include example code for various illustrations of environmental data
Specific Topics-- After the initial Learn R course, I expect to host shorter sessions with fewer participants on topics such as analysis of species composition data, generating GRTS and related samples, etc.. In order to capture that information for later use, I'll try to build pages here.
General Information about R
Developers of R describe it as a programming language and environment for data analysis and graphics (Venables, Smith, & Core R Team 2005). It's been used to develop an integrated software suite that can be used to conduct an exceptionally broad range of statistical analyses, and the entire package is open source software that is freely available to the public. The R language is a object-oriented and highly extensible, thus it's the development language of choice for many statisticians that are developing or implementing new techniques and procedures. As a result, newer statistical procedures are available in R long before they appear in other statistical software packages.
For our purposes, R consists of a core package and a very large number of "packages". The standard download will install software that provides all the functionality that many users will ever need, but you will periodically want to check for updates and enhancements to core software. Contributed packages are posted to the R project web sites for examination and downloading. R makes it very easy to update or install packages - just use the 'update' or 'install' commands under the "Packages" main menu selection.
An R GUI (Graphical User Interface)
In the past, the absence of a graphical user interface (GUI) was a significant impediment to using R. Now there are reasonable alternatives. The base R distribution includes a simple GUI that provides for core program functions, such as automatically installing and updating packages. For conducting analyses, Rcmdr is a GUI that is sufficiently stable and feature-rich to merit recommendation. Rcmdr is a GUI that handles many routine analyses - regression, ANOVA, categorical analyses, cluster analyis, etc. Development and enhancement of the GUI is an active area, and updates are posted regularly on the respective web sites.Basic Resources for R |
|
| R project home | Core R project page. Main R site -for all operating systems and resources. Use next link to obtain current Windows version of R package. |
| R for Windows | Download latest Windows version via this link. |
| Rcmdr package | R GUI developed by John Fox. Handles a variety of routine parametric and non-parametric tests, and facilitates production of a range of graphics. Makes R accessible to the casual user. This is really cool. |
| TinnR | R GUI, text editor, and some other stuff. If you try these, please send me a review. The R project home has a link to other GUIs. |
Compendium: R for Analysis of Ecological and Environmental Data
Click here to view the R community's compendium / review / synthesis of packages that are generally used for ecological analyses. Includes species modeling, ordination, time series, TREE, extreme value analysis, and related topics. An excellent resource. There is also a new R listserve specifically for ecological and environmental analysis.
Getting Help in R
Aside from the documentation available on the CRAN website, there are at least 2 tools that are better than a raw google search. My first step is often to go to cran
http://cran.r-project.org/
select "packages", then search in the webpage for the text I want (e.g., Oracle or SAS).
Second, I have Jonathan Baron's R search page bookmarked:
http://finzi.psych.upenn.edu/search.html
It searches most of the R online help, plus the archives of the R-help mailing lists.
Many universities offer courses on R; several offer courses on using R in natural resources. UCLA ATS has a growing set of R (and other statistical computing) pages at:
The national Center for Ecological ANalysis and Synthesis (nceas.ucsb.edu) offers no-nonsense training in R for ecologists with strong quantitative and programming background. Unfortunately, their course web pages which were a wonderful resource are no longer visible.