Category Archive

We’re very pleased to announce the release of devtools 1.0.  We’ve given devtools the 1.0 marker because it now works with the vast majority of packages in the wild, with this version adding support for S4 and Rcpp.  Devtools also has completely revamped code for finding Rtools on windows, including much better error messages if something is wrong with your setup. In celebration of reaching 1.0, devtools now has it’s own webpage, which provides a bit more information about why you might want to use it.

Devtools 1.0 also contains many other bug fixes and minor improvements, as listed in the NEWS file on github.

Earlier this month a new version of the Rcpp package by Dirk Eddelbuettel and Romain François  was released to CRAN and today we’re excited to announce a new version of RStudio that integrates tightly with Rcpp.

First though more about some exciting new features in Rcpp 0.10.1. This release includes Rcpp attributes, which are simple annotations that you add to C++ source files to streamline calling C++ from R.  This makes it possible to write C++ functions and simply source them into R just as you’d source an R script. Here’s an example:

#include <Rcpp.h>
using namespace Rcpp;

// [[Rcpp::export]]
NumericMatrix gibbs(int N, int thin) {

   NumericMatrix mat(N, 2);
   double x = 0, y = 0;

   RNGScope scope;
   for(int i = 0; i < N; i++) {
      for(int j = 0; j < thin; j++) {
         x = R::rgamma(3.0, 1.0 / (y * y + 4));
         y = R::rnorm(1.0 / (x + 1), 1.0 / sqrt(2 * x + 2));
      }
      mat(i, 0) = x;
      mat(i, 1) = y;
   }

   return(mat);
}

By annotating the gibbs function with the Rcpp::export attribute, we indicate we’d like that function to be callable from R. As a result we can now call the function like this:

sourceCpp("gibbs.cpp")
gibbs(100, 10)

Thanks to the abstractions provided by Rcpp, the code implementing gibbs in C++ is nearly identical to the code you’d write in R, but runs 20 times faster.

The sourceCpp function makes it much easier to use C++ within interactive R sessions. In the new version of RStudio we did a few things to support this workflow. Here’s a screenshot showing the RStudio C++ editing mode:

In RStudio you can now source a C++ file in the same way as an R script, using the source button on the toolbar or Cmd+Shift+Enter. If errors occur during compilation then RStudio parses the GCC error log and presents the errors as a navigable list.

When using sourceCpp it’s also possible to embed R code within a C++ source file using a special block comment. RStudio treats this code as an R code chunk (similar to Sweave or R Markdown code chunks):

RStudio also includes extensive support for package development with Rcpp. For more details see the Using Rcpp with RStudio document on our website.

Note that if you want to try out the new features be sure you are running RStudio v0.97.237 as well as the very latest version of Rcpp (0.10.1) .

Today a new version of RStudio (v0.97) is available for download from our website.  The principal focus of this release was creating comprehensive tools for R package development. We also implemented many other frequently requested enhancements including a new Vim editing mode and a much improved Find and Replace pane. Here’s a summary of what’s new in the release:

Package Development

• A new Build tab with package development commands and a view of build output and errors
• Build and Reload command that rebuilds the package and reloads it in a fresh R session
• Create a new package using existing source files via New Project
• R documentation tools including previewing, spell-checking, and Roxygen aware editing
• Integration with devtools package development functions
• Support for Rcpp including syntax highlighting for C/C++ and gcc error navigation

Source Editor

• Vim editing mode
• Tomorrow suite of editor themes
• Find and replace: incremental search, find/replace in selection, and backwards find
• Auto-indenting: improved intelligence and new options to customize indenting behavior
• New options: show whitespace, show indent guides, non-blinking cursor, focus console after executing code

More

• New Restart R and Terminate R commands
• More intelligent console history navigation with up/down arrow keys
• View plots within a separate window/monitor.
• Ability to set a global UI zoom-level
• RStudio CRAN mirror (via Amazon CloudFront) for fast package downloads

There are also many more small improvements and bug fixes. Check out the v0.97 release notes for details on all of the changes.

At RStudio, we want you to be effective R users. As well as creating great software, we want to make it easier for you to master R. To this end, we’re very happy to announce our new training offerings.

We’re kicking off with two public courses:

• Effective data visualisation and reports and reproducible research in San Francisco, Dec 3-4.
• Advanced R programming and package development in Washington DC, Dec 10-12.

We’ve also planned a number of other courses, based on our experience with the R community, seeing what’s hard to learn and what people are struggling with. These courses are available now if you’d like us to come to your company, and based on your feedback we’ll offer public versions in the near future.

You can also read about our instructors and our philosophy.

RStudio’s mission from the beginning has been to create powerful tools that support the practices and techniques required for creating trustworthy, high quality analysis. For many years Hadley Wickham has been teaching and working on his own set of tools for R with many of the same core goals. We’ve been collaborating quite a bit with Hadley over the past couple of years and today we’re excited to announce that Hadley, Winston Chang, and Garrett Grolemund are joining RStudio so we can continue to work together much more closely.

You probably know Hadley from his work on ggplot2, plyr, and many other packages. Garrett was a PhD student of Hadley’s at Rice, and you might also know him from the lubridate package, which makes dealing with dates and time easier; he’s also been working on new tools for visualisation and new ways of thinking about the process of data analysis. Winston has been working full-time on ggplot2 for the last couple of months, squashing many bugs and repaying a lot of the technical debt that’s accumulated over the years. Winston’s also writing an R Graphics Cookbook for O’Reilly that should be available in the near future.

What does this mean for RStudio? We’ll of course continue developing open-source software like the RStudio IDE, ggplot2, and plyr (among many other projects). One of Hadley’s core focuses at RStudio will also be expanding our mission to include education, which we plan to offer in a variety of formats ranging from in-person training to some innovative new online courses. We’ll also be working on hosted services (like RPubs) as well as some new products that address the challenges of deploying R within larger organizations.

We’re all excited to begin this next phase of work together and will have lots more details to announce later this fall!

Today we’re very excited to announce RPubs, a free service that makes it easy to publish documents to the web from R. RPubs is a quick and easy way to disseminate data analysis and R code and do ad-hoc collaboration with peers.

RPubs documents are based on R Markdown, a new feature of knitr 0.5 and RStudio 0.96. To publish to RPubs within RStudio, you simply create an R Markdown document then click the Publish button within the HTML Preview window:

RPubs documents include a moderated comment stream for feedback and dialog with readers, and can be updated with changes by publishing again from within RStudio.

Note that you’ll only see the Publish button if you update to the latest version of RStudio (v0.96.230, available for download today).

The markdown package

RStudio has integrated support for working with R Markdown and publishing to RPubs, but we also want to make sure that no matter what tools you use it’s still possible to get the same results. To that end we’ve also been working on a new version of the markdown package (v0.5, available now on CRAN).

The markdown package provides a standalone implementation of R Markdown rendering that can be integrated with other editors and IDEs. The package includes a function to upload to RPubs, but is also flexible enough to support lots of other web publishing scenarios. We’ve been working with Jeff Horner on this and he has a more detailed write-up on the capabilities of the markdown package on his blog.

Gallery of examples

We’ve also published a gallery of example documents on RPubs—the gallery illustrates some of the most useful techniques for getting the most out of R Markdown, and includes the following articles:

• MathJax and Writing Equations
• Dynamic Graphics with the googleVis Package
• Customizing Chunk Options
• Caching Code Chunks

Let us know what additional examples you’d like to see—we’ll be adding more in the weeks ahead.

We’ve just a made a change to the syntax for embedding MathJax equations in R Markdown documents. The change was made to eliminate some parsing ambiguities and to support future extensibility to additional formats.

The revised syntax adds a “latex” qualifier to the $ or $$ equation begin delimiter. It looks like this:

This change was the result of a few considerations:

1. Some users encountered situations where the $<equation>$ syntax recognized standard text as an equation. There was an escape sequence (\$) to avoid this but for users not explicitly aware of MathJax semantics this was too hard to discover.
2. The requirement to have no space between equation delimiters ($) and the equation body (intended to reduce parsing ambiguity) was also confusing for users.
3. We want to eventually support ASCIIMath and for this will require an additional qualifier to indicate the equation format.

RStudio v0.96.227 implements the new MathJax syntax and is available for download now.