-
-
-
-
-
- Assignment
-An object is an entity that contains information and can be manipulated by commands. R has two main commands for assigning an object: ‘\(<\)-’ and ‘=’.
-> x <- 5
-> x = 5
-We will use ‘=’ throughout this document. However, many R users prefer ‘<-’, because ‘=’ is used for other things, too. A third method is very rarely used:
-> 5 -> x
-Each of the previous commands assigns the number 5 to the object x. Notice that R produces no output when the above commands are run. In order to see what R has done, type:
-> ls()
-and/or look at the environment window in the upper right corner. Now type
-> x
-When you submit a command to R, one of three things can happen.
-
-You see a result: e.g.,
-> x
-R prints the value of the expression.
You see nothing except another command prompt: e.g.,
-> y = log(x)
-For an assignment, R stores the value of log(x) in the object y, but produces no output.
You see an error message: e.g.,
-> y = lg(x)
-Look at error messages – they can be informative!
-
-
- Manipulating Objects
-We can perform mathematical operations on objects such as x.
-> x + 2
-Notice that x has not changed:
-> x
-We can change the value of x:
-> x = x + 2
-> x
-Cautionary Tip: It is very important to use caution when writing over a variable as above. If you need to use x later on, be sure you are using the correct value!
-Start from scratch and perform operations on two objects.
-> x = 5
-> y = 2
-> x - y
-If two objects are assigned to have the same value, they can be changed to differ. (Assigned by value not by assigned by reference, for those of you who know what that means.)
-> a = 3
-> b = a # Note: Assignment
-> b == a # Note: Test of Equality
-> a = a + 1
-> a
-The value of b didn’t change.
-> b
-Assign a vector of numbers to the object x
-> x = c(3, 5, 9, 10)
-> x
-Get a list of the objects in the workspace.
-> ls()
-Remove an object.
-> rm(x)
-
-
- Indexing Objects
-Situations frequently arise when you want to access select portions of a database. In this section, we discuss how to extract elements of vectors and matrices.
-
- Indexing Vectors
-> x = c(13,21,99,10,0,-6)
-Suppose that we only need the first element of the vector x. To extract the first element, we type the name of the entire vector, followed by the index we want to extract enclosed in brackets.
-> x[1]
-We can save the extracted part to a new object
-> z = x[1]
-> z
-We often will want to extract more than one element of a vector. Each of the following two lines of code extracts the first three elements of the vector x.
-> x[c(1,2,3)]
-> x[1:3]
-What happens if we try to extract the first three elements in the following way?
-> x[1,2,3]
-Elements can be extracted in any order and elements can be extracted any number of times. All of the following are legitimate methods of extracting multiple elements from a vector.
-> x[c(2,4,5)]
-> x[c(4,5,1)]
-> x[c(5,1,5,2,1,1,1,5)]
-The following code extracts all elements of x except the second.
-> x[-2]
-What will this do?
-> x[-c(2,4)]
-
-
- Indexing Matrices
-To extract an element from a matrix, you may specify two values: the row value and the column value. The row and column are separated by a column.
-> M1 = matrix(1:12, nrow=3, byrow=TRUE) # (this is obj5 from before, so M1 = obj5 works too)
-> M1
-Pick out the number from the second row and third column.
-> M1[2,3]
-You can simultaneously select multiple rows and multiple columns.
-> M1[2,c(1,3)]
-> M1[c(2,3),c(1,2)]
-If nothing is specified in the row position (before the comma), then every row is kept. Similarly, every column is kept if nothing is specified in the column position.
-> M1[,c(2,3)]
-> M1[c(1,2),]
-If nothing is specified in either position, the entire matrix is returned.
-
-
-
- Index Assignment
-In addition to extracting certain indices, it is also possible to assign new values to certain elements of a vector or matrix.
-The following two lines of code change an element of the vector x and the matrix M1.
-> x[3] = 5
-> M1[2,3] = 6
-
-
- Aside: Missing Index?
-If an index is missing, it might be any index. This is rarely what you want: Avoid missing values in your index.
-> x[NA]
-
-
- Object Classes
-So far we seem to have been working exclusively with numeric objects. R can store objects of many different types. Suppose you are working with a data set that includes both quantitative and categorical variables. R can store these as different classes. Let’s begin by looking at two basic classes, numeric and character.
-> x = 12
-> class(x)
-> y = c(3,5,2)
-> class(y)
-R stores both the number 12 and the vector c(3,5,2) as an object of the class numeric. Strings are stored as characters.
-> x = "Hi"
-> class(x)
-> y = c("sample", "string")
-> class(y)
-Elements of vectors and matrices must be of the same class.
-> mix = c("aa", -2)
-> mix
-> class(mix)
-> mix[2]
-> class(mix[2])
-When working with data, this will create problems if a column representing a quantitative variable contains character text. The numeric is promoted to character.
-
-
- How to Mix Variables of Different Classes
-Matrices are not well-suited for storing data sets. Data sets frequently contain different types of variables (quantitative, qualitative). Matrices force all elements to be of the same class. A data.frame is particularly adept at handling data of different classes.
-> num = c(2,9,6,5)
-> char = LETTERS[c(24,24:26)]
-> dat = data.frame(num, char, stringsAsFactors=FALSE)
-> dat
-> class(dat)
-Though data analysts will rarely spend their time investigating a data set as small this one, exploring data sets such as these can be helpful in learning R’s capabilities. In the following code, we investigate the names and dimensions of the data set dat; we also investigate the properties of the columns of dat.
-> names(dat)
-> dim(dat)
-> nrow(dat)
-> ncol(dat)
-> class(dat[,1])
-> class(dat[,2])
-R stores the first column as numeric and the second column as a character. summary gives a numerical summary of numeric variables and little useful information for character variables.
-> summary(dat)
-It is likely that you want to store a categorical variable as a factor rather than a character vector. The default behavior of data.frame to do the conversion.
-> dat = data.frame(num, fac=char)
-Now R stores the first column as numeric and the second column as a factor. summary gives a numerical summary of numeric variables and a table for categorical variables.
-> class(dat[,2])
-> summary(dat)
-Keeping track of column numbers can be tedious. It is often more convenient and cleaner to index by the column name. Name indexing uses the dollar sign ($) or double square braces ([[]]).
-> dat$num # Or dat[["num"]]
-> dat$fac # Or dat[["fac"]]
-Factors can be created explicitly (not just as a side effect of the data.frame function)
-> fac = factor(char)
-The levels function returns the levels of a factor.
-> levels(fac)
-
-
-
-
-
3.2 Comments
-Documentation and formatting are essential to writing effective R code. If you come back to a project after a few months (days? hours?), you want to know what the code is doing without retracing every single step. Comments in R can be inserted with the
-#symbol. R will not process the rest of the line after the#.
-> 5 < 6
-> 5 # < 6The following is an example of a commented R script. Some of the functions we have used before; others will be explained later. Let’s open it in RStudio and see what it does.
-