Skills for R

Posted on Nov 14, 2012 in Programming

Things under legendu.net/outdated are outdated technologies that the author does not plan to update any more. Please look for better alternatives.

Unlike C/C++, Java and SAS, you do not have semicolon to indicate the end of a command, but you can do this if you like. It is helpful if you have to write multiple commands on a line.
R support command/terminal mode, and if you open R in a command/terminal window, the working directory of R will be the working directory of the command/terminal window. For some GUI for R, you can only use a single application, e.g. you cannot open two Rstudio at the same time. But sometimes, you might want to open more than one R applications in different directories. To do this, you can use the command/terminal mode.
When you do not know how use a function, you can always type a question mark before the function name to open the help document of the function. For example, if you do not know how to use sum, you can type in ?sum in R console to open the help document for sum. Sometimes, you might just want see what arguments a function have, instead of opening the help document. In this case, you can use args which display the declaration of a function. It is quite often that you want to use a function in R, but you cannot remember the name of the function. If you still know part of the function name, you can type in ??partial_name in R console, then R will try to match it with all possible functions using regular expression.
Sometimes, you might want to hide an object (i.e. do not display the object when using the command ls()) in the R workspace, for example if you write a package which relies on some global variables or some functions that should never by renamed. To do this, you can give a name to the object starting with a dot. For example there are some functions such as .jinit, .jcall and so on in R rJava, but you cannot see them by use ls(). If you do want to display all objects including these that start with a dot in R workspace, you can use ls(all.names=TRUE).
By default, numerical variables in R are double values even you even they do not have decimal digits. To use data type of integer, you can put an L after integers. For example to assign a integer 5 to a variable aInt, you can use aInt = 5L.
comment can be use to set or query comment of an object in R.
with allows you evaluate an expression inside a data environment (usually a data frame), so you do not have to attach the data into R workspace (Attaching data into R workspace might mark other variables with the same names, so it is encouraged to do so.). within transform
class gets/sets the class of a R object.
attributes and attr can get and set attributes of objects, e.g. it can list names of objects. The difference between them is that attributes can set the attributes of a null object.
str shows the structure of an arbitrary R object. For example, if you have a large data frame df you can use str(df) to show its structure, so that you know what columes (type of data and names) it has.
The ESC hot key halts current unfinished command, which is helpful when you make typo in a command (especially when you miss a single/double quotation mark).
Rprof enables profiling an R is execution, which helps you write efficient R code.
If you want to run some code without writing any intermediate result into the workspace, you can put the code into local surrounded by curly braces. However, you must use -\> or <- instead of = in the code. Note that you can use semicolon to separate different commands.
exists checks whether a R object with a given name has been defined. For example,

> exists("aNon_Exist_Object")
[1] FALSE

missing checks whether an argument of a function has been passed a value or not. It can only be used inside a function to help check availability of arguments.

You can use shell("shutdown/s") (in Windows) to shutdown the computer after R has finished computing.
When write a user-defined function, you'd better return a comprehensive result which contains not only the final outcomes but also other information such as parameters used in the computation. This gives you the ability to work on a project in the long term, especially when the names of variables do not give you much information about parameters used in computation (typical they do not because otherwise they will be long).