R does not, at this time, offer interactive (dynamic) graphics, in contrast to or other software for high interaction statistical work like or , S-plus has a few truly interactive functions, namely the brush() and spin() functions.

R has a few, rather basic functions namely the identify() and locator() functions. (details here).

Note that there is a lot of development going on in this area, the information below is certainly incomplete and likely to be quickly outdated

There are several packages that offer interactive graphics for their specific purpose, they often require the installation of additional libraries inside and software libraries outside R, but these packages are self-contained, as R has not been conceived for interactive graphics. Examples of this are:

• Package playwith (it requires the installation of GTK (a cross-platform widget toolkit for creating graphical user interfaces).
• Package aplpack adds more graphics and some interactive slider functions.

Other packages rely on web-based technologies (html5, javascripit) using a browser for easy sharing, display and interaction. Most packages, when run within Rstudio, display the interactive charts in the Rstudio viewer window. Examples are:

• Package ggvis lets you add sliders and other interactive functions (displayed an a browser window) [ggvis]
• Package googleVis produces google motion charts with r (have a look at demo(googleVis)
¨
• Package (related to the plotly chart sharing web-site), displays its charts in a browser window or Rstudio viewer window
li>clickme Scatterplot. Sharable dynamic plots; templates for different types [I never succeeded to install it correctly on Windows; devlopment seems slow (project abandonned?] [Website with some examples]
• The animint converts ggplot2 graphics to javascript to provide interaction.
• cranvas is an R package built on Qt libraries and provides interaction for various charts.

The current path to highly interactive data analysis uses R together with other software.

• The currently most mature approach is to use GGobi, highly interactive open source software for multidimensional data exploration. There is a full bidirectional interface between the two software systems; R functions can pass control to GGobi for particular analysis tasks, and vice-versa.
• As more and more commercial software is able to interact with R, the specific interactive facilities of software such as or are wide open for R users.

There is also more and more software that has a data interface with R, i.e. it uses R to provide data and some additional facilities. Two interesting examples of this are and .

Several software projects are underway to provide a more complete graphical user interface to R and adding interactive graphics to R. Be aware that the native graphics system of R is not affected by this, i.e. to get the interaction you have to use the specific functions provided.

A good example of this is the package iplots, integrated into JGR ("Jaguar-R) a Java based alternative user interface. More recently the Deducer project has added a menu driven analysis interface to the JGR environment[Deducer].

Rstudio manipulate package lets you modify graphical options of a particular chart.

• Shiny A web application framework for R (Rstudio)
<