R workspace, object and .RData file

The R program’s structure is similar to the programs written in other computer languages such as C or its successors C++ and Java. However, important differences between these languages and R are (i) R has no header files, (ii) most of the declarations are implicit, (iii) there are no pointers in R, and (iv) text and strings as vectors can be defined and manipulated directly.

R is a functional language. Most of the computation in R is handled using functions. The R language environment is designed to facilitate the development of new scientific computation tools.

Every thing (such as functions and data structure) in R is an object. Too see the names of all objects in R workspace, on R command prompt just type,


objects() is an alternative to ls() function. Similarly, typing the name of any object on R prompt displays (prints) the content of that object. As an example type q, mean, and lm etc. on R prompt.

It is possible to save individual object or collection of objects into a named image file. The named image file have extension of .RData. Som e possibilities to save object from R workspace are:

To save content of R workspace into a file .RData, type

> save.image()

To save objects in file archive.RData, type

> save.image(file = “archive.RData”)

To save some required objects in data.RData, type

> save(x, y, file = “data.RData”)

These image files can be attached to make objects available in the next R session. For example.

> attached (“arvhive.RData”)

R workspace

Note that when quitting, R offers the option of saving workspace image. By default workspace is saved in an image file (.RData) in working directory. The image file can e used in the next R session. Saving the workspace image will save everything from current workspace. Therefore, use rm() function to remove objects that are not further required in next R session.

For further details about saving and loading R workspace visit: http://rfaqs.com/saving-and-loading-r-workspace

Leave a Reply

%d bloggers like this: