RStudio Beta 2 (v0.93) is available for download today. We’ve gotten incredibly helpful input from the R community and this release reflects a lot of that feedback.
The release notes have the full details on what’s new. Some of the highlights include:
Source Editor Enhancements
- Highlight all instances of selected text
- Insert spaces for tabs (soft-tabs)
- Customizable print margin line
- Selected line highlight
- Toggle line numbers on/off
- Optional soft-wrapping for R source files
Customizable Layout and Appearance
- The layout of panes and tabs is now configurable (enabling side-by-side source and console view, among others).
- Support for a variety of editing themes, including TextMate, Eclipse, and others.
Interactive Plotting
This release features manipulate, a new interactive plotting feature that enables you to create plots with inputs bound to custom controls (e.g. slider, picker, etc.) rather than hard-coded to a single value. For example:
manipulate(
# plot expression
plot(cars, xlim = c(0, x.max), type = type, ann = label),
# controls
x.max = slider(10, 25, step = 5, initial = 25),
type = picker("Points" = "p", "Line" = "l", "Step" = "s"),
label = checkbox(TRUE, "Draw Labels")
)
More
- RStudio now works with versions of R installed from source (either via make install or packaged by MacPorts, Homebrew, etc.).
- Enhanced support for Unicode and non-ASCII character encodings.
- Improved working directory management including new options for default behavior, support for shell “open with” context menus, and optional file assocations for common R file types (.RData, .R, .Rnw).
- Many other small enhancements and bug fixes (see the release notes for full details).
We hope you try out the new release and keep talking to us on our support forum about what works, what doesn’t, and what else you’d like RStudio to do.





32 comments
Comments feed for this article
April 11, 2011 at 1:19 pm
teucer
First of all, congratulations for the fantastic job!The great progress is really visible.
However, I would like to share some observations:
*I don’t understand the syntax coloring of seq(-pi/2,3*pi/2,length.out=25): pi/2,3*pi/2 is interpreted as a string
*workspace is not working properly in a “browser” environment: variables are not listed, tab-completion is not working
*for dual screens it would be nicer to have a separate plot window
April 11, 2011 at 3:12 pm
jjallaire
Hi there,
Thanks for the bug report and feedback. We’ll note all of these as defects and hopefully improve upon them soon.
J.J.
April 12, 2011 at 5:07 am
Paolo
GREAT Job in general and good new version: the editing themes and the easy set up of the default working directory are my favorite new features! Now The only feature I miss is the Column Mode (rectangular selection)! Keep up with the good work!
April 12, 2011 at 5:09 am
nico
A quick and dirty wrapper to call manipulate and automagically have manipulators for x and y
Hope it can be helpful for someone.
Just use it exactly as a normal plot call.
plot.manip <- function(x, y=NULL, …)
{
if (is.null(y))
{
y <- x
x <- NULL
xlim <- c(1,length(y))
ylim <- range(pretty(y*1.1))
xstep <- diff(xlim)/100
ystep <- diff(ylim)/100
manipulate(
plot(y, xlim=c(x.min, x.max), ylim=c(y.min, y.max), …),
x.min = slider(xlim[1], xlim[2], step = xstep, initial = xlim[1]),
x.max = slider(xlim[1], xlim[2], step = xstep, initial = xlim[2]),
y.min = slider(ylim[1], ylim[2], step = ystep, initial = ylim[1]),
y.max = slider(ylim[1], ylim[2], step = ystep, initial = ylim[2])
)
}
else
{
xlim <- range(pretty(x))
ylim <- range(pretty(y*1.1))
xstep <- diff(xlim)/100
ystep <- diff(ylim)/100
manipulate(
plot(x, y, xlim=c(x.min, x.max), ylim=c(y.min, y.max), …),
x.min = slider(xlim[1], xlim[2], step = xstep, initial = xlim[1]),
x.max = slider(xlim[1], xlim[2], step = xstep, initial = xlim[2]),
y.min = slider(ylim[1], ylim[2], step = ystep, initial = ylim[1]),
y.max = slider(ylim[1], ylim[2], step = ystep, initial = ylim[2])
)
}
}
April 12, 2011 at 9:23 am
Steve Lianoglou
Impressive list of new features. Keep up the good work!
April 12, 2011 at 10:16 am
G.
Hi,
first congratulations for this great job. Maybe it is not the right topic but here are my observations (let’s say I work with the default layout):
- We cannot choose the sorting option in a folder when exploring it in the lower right pane (the default option is “by name”)
- in the plot pane (still lower right), when I use the “export” option, a new panel appears with the plot on it. But when I double click on the image, the background colour changes and all the other parts of the panel can be selected.
- Are you planning to introduce a synchronization between the variables and functions values in the code and in the workspace? In the current version, if you double click on a function in the upper right pane (Workspace), you can modify it in the workspace but with no effect on the code. Such an option (like “attach this file to the current workspace”) could also lead to a “variable view” option in the code.
Goulven
April 12, 2011 at 11:45 am
jjallaire
Hi there, in response to your questions:
- Yes, we need to add sorting options for the files pane.
- The wierd selection behavior you observed is due to the fact that we are using webkit as our UI engine (so we inherit the standard selection behavior of a webpage). Over time we hope we can eliminate these cases.
- For the synchronization behavior you suggested, you could try using the “Source on Save” option in the source editor (which automatically sources the file into the environment every time you save it, thus ensuring that the code and workspace stay in sync).
April 12, 2011 at 12:42 pm
G.
Thanks for the quick answers. I didn’t notice the “Source on Save” option. Still, when editing a function from the workspace pane (which can be pretty convenient to locate it in a large code), there is no effect on the actual code (the .R file). That is why I mentioned another option.
Thank you
G.
April 12, 2011 at 11:35 am
Paolo
That’s great.
What’s the best way to upgrade?
Uninstalling previous version and reinstalling or simply install new version onto previous one?
Paolo
April 12, 2011 at 11:42 am
jjallaire
Paolo,
Yes, just re-run the installer and it will perform the upgrade.
J.J.
April 15, 2011 at 7:53 am
Eric Nord
After using it a few days, I really like it. The best GUI for R on Mac or Windows that I have seen.
I love the autocompletion and the wonderful function hinting/help integrated with the autocompletion.
I’d like to see greater flexibility in setting up the panes, eg ability to have 2 panes dedicated to source.
Another really cool feature would be some way to embed a simple web browser in a pane so you could have an “Rseek” pane.
Thanks! I’ll be recommending this.
April 18, 2011 at 12:01 am
Memo Garza
Just want to register my compliments. It’s freaking beautiful. It loads fast and works great. I would make a few comments, but I think at the pace you’re going you’ll surpass anything I could suggest.
April 18, 2011 at 11:25 am
Paolo
It would be great to have the edit() functionality in place
So far I get the following…
“Editing of matrix and data.frame objects is not currently supported in RStudio.”
Paolo
April 18, 2011 at 10:07 pm
Mario Morales
I’m using Studio and is excellent. However I would like to export the graphics to formats such as ps and eps.
April 19, 2011 at 6:11 am
jjallaire
Mario,
Thanks for your comment — enhanced image export options is on our short list so you should see some improvements in this area in an upcoming release.
J.J.
April 21, 2011 at 6:31 am
Paolo
I’m using RStudio from an external USB drive which is great.
Btw I have just noticed that when I open a new R script and don’t save it, the next time I open Rstudio it is still there only when opening from the same computer…is it possible to fix it?
Paolo
April 23, 2011 at 11:49 am
jjallaire
Paolo,
Yes, this is a limitation of the current version of RStudio. If you are on Windows perhaps there is a way to redirect AppData\Local (which is where we store the unsaved scripts) to your USB key?
J.J.
May 2, 2011 at 2:10 pm
Nice
I’ve been using a Linux terminal + gedit for years… I was never convinced that the IDEs for R were polished enough (they always seemed too buggy)… RStudio may have just changed my mind. Looks great so far, keep up the good work.
May 3, 2011 at 1:35 pm
Paolo
GRAZIE!!!
GREAT GUI
May 6, 2011 at 8:47 pm
aaa001
Hi,
Is it a good idea to add a button on an opened R scrip file to change the working directory to the current file location?
Regards
aaa001
May 6, 2011 at 9:36 pm
jjallaire
Yes, we’ve heard from a few people that adding some more gestures for setting the working directory would be beneficial and we’re hoping to improve on this in the next release.
J.J.
May 17, 2011 at 10:34 am
jeff
that’s a great job what you did there!
R really needed that. I used to programm with tinn-R. Both have their pros and cons but i preffer rstudio because i like the panes layout and the way the graphs are stored and the fact that the console code is being colored.
I use R evryday, that’s the tool i’m working with for my job and there is still some kind of improvements that i hope you’ll make and which i used to appreciate with tinn-R.
1) to be able to customize the shortcuts (especially to run lines code)
2) to have the possibility to select end replace character in just the selected code
3) to have the possibility to click on next when you want to find or replace some character
4) to highlight some special programming characters like : for, if, function, plot etc…in other colours to make the code even easier to read
I really thank you for making this essential software.
see you,
j-f
May 22, 2011 at 11:58 am
John
Fantastic job, all. RStudio is amazing
May 28, 2011 at 12:56 am
zhanxw
Great job! I am recommending RStudio to every one around!
June 2, 2011 at 12:13 am
fRed
You should have a bug and improvement list like the google code framework. Anybody could add / edit bugs and improvement and it would help you prioritize!
Otherwise, thanks, I am still using Eclipse with Statet but I am considering moving to RStudio if it continues to improve at such a great pace..
June 2, 2011 at 5:23 am
jjallaire
Thanks, yes, we have our support forum (http://support.rstudio.org) however it doesn’t have the editing and voting features we’d like it to. Something we’re definitely hoping to do. Stay tuned lots more good stuff coming…
J.J.
June 3, 2011 at 6:01 am
Kenn
Great program! But I’m having trouble with workspaces and working directories. My favorite way of working with R is keeping everything related to a project in a separate folder; when I double click the .Rdata file then the working directory is automatically set to that folder, the related history is automatically available etc. With Rstudio I can’t count on that. The history seems to be global i.e. when I open project1, close it and open project2, then the history will be that of project1 (and possibly preceded by something unrelated). Saving a script file offers some unrelated folder by default. Finally, I can’t figure out where is the workspace saved when I click “save” when closing R+RStudio. It works as expected at times, but then at other times, after making several computations and “saving” a workspace, I find that nothing was changed in the working directory (i.e. the .Rdata file still contains only what was there before working with it with RStudio). This prevents me of being a big fan of Rstudio yet.
June 3, 2011 at 1:37 pm
jjallaire
Kenn,
Agreed that the current iteration of the product doesn’t handle this very gracefully. In our upcoming beta release there will be some improvements (mostly aimed at making us behave identically to standard R with respect to working directories):
- History will be per-working directory (although the user can still search history globally as well as override the per-working directory behavior if they prefer the global approach)
- Open File and Save As will default to the current working directory
With respect to the workspace seemingly not being saved — we will save the .RData and .Rhistory file to the working directory at the time when you exit R. This is the same behavior as console R and R GUI, although I agree that this has the potential for creating confusion and seemingly “lost” working context. Our longer term solution to this will be a project system that preserves the working directory methodology you outlines but makes stronger guarantees about what will be loaded and saved (the project directory will determine where things are saved rather than whatever happens to be the working directory when you exit R).
If you want to try out some of these features now (the history is the main one currently implemented) you can check out our daily build at:
http://www.rstudio.org/download/daily/desktop/
Thanks for the feedback and I hope we are able to address all of your concerns in our upcoming releases.
J.J.
June 6, 2011 at 5:06 am
Kenn
Thanks! I’ll try it out …
If RStudio’s behavior will be organized around “projects” then maybe an option to “save project” would be good. This could be just a very small text file naming the workspace and history file, as well as the opened script file(s), and maybe also screen layout. Opening this “Rstudio project file” would then restore the session as it was when Rstudio was last closed with this project.
Another little feature I’ve thought of is shortcuts for changing the screen layout quickly. For example, on my 12″ laptop, I can’t keep all 4 screen areas of equal size; when switching from code to console and back, I usually “almost maximize” the area where I’ll be working. It would be very convenient if this could be done with a keyboard shorcut, for example, Ctrl-Shift-2 would switch focus to the console, and “almost maximize” (depending on some user settings) its area, and Ctrl-Shift-1 would do the same for the source code area.
Thank you for a highly useful piece of software!
June 6, 2011 at 6:14 am
jjallaire
Kenn,
Yes, the project feature will very be very much along the lines of what you described.
Good idea on the switch and maximize feature — we’ll definteily consider that!
J.J.
June 5, 2011 at 7:42 pm
carol
Dear friends,
I love RStudio!
I hope you can add more functionality in the future. Something I would like to see is little checkboxes in the workspace objects listing that I could click, then “delete” or even save as object_name.Rdata.
Thanks!
Carol
June 5, 2011 at 8:13 pm
jjallaire
Carol,
Thanks for the kinds words. Yes, we’re adding new functionality all the time (you can expect another beta release soon). We’ve definitely targeted the workspace as an area for lots of future improvement.
J.J.