What is R and RStudio
R is a software environment for performing statistical analyses. RStudio is a helper program that makes it easier to use R. Below are four videos to introduce you to R and RStudio.
- What is R?
- What’s Up With R? [stop at 1:51]
- What is RStudio (part 1) [stop at 2:27]
- What is RStudio (part 2) [only from 3:26 to 4:13]
Detailed methods for downloading, installing, and configuring R, RStudio, and NCStats on your personal computer are given on the Resources page of the course website.
Installing R, RStudio, NCStats
R and RStudio can be installed to your personal computer, but does not work well if you have a ChromeBook or if your machine is old (especially Macs). If you have a ChromeBook or old machine then please follow the “RStudio-in-the-Cloud” directions further below. Otherwise follow the Windows or Mac directions below.
Windows
To install R on a Windows machine, download the installer from this site by selecting the “Download R X.X.X for Windows” link at the top. Locate the installer on your computer and run it. Using all of the default choices will perform an acceptable install for this class.
To install RStudio on a Windows machine, download the installer from this link by selecting the “DOWNLOAD RSTUDIO FOR WINDOWS” button. Locate the installer on your computer and run it. Using all of the default choices will perform an acceptable install for this class.
Install NCStats as described here.
Mac
To install R on a Mac, download the installer from this site by selecting one of the “R-X.X.X.pkg” links appropriate for your operating system under the “Latest Release” section. Locate the “.pkg” file on your computer and install it per usual. Using all of the default choices will perform an acceptable install for this class.
To install RStudio on a Mac, download the installer from this link by selecting the “DOWNLOAD RSTUDIO FOR MACOS” button (you can also select the MacOS option under the “All Installers” section). Locate the “.dmg” file on your computer and install it per usual. Using all of the default choices will perform an acceptable install for this class.
Install NCStats as described here.
NCStats
The latest NCStats may be installed by running these lines in R/RStudio (i.e., copy these three lines from below, paste them into the CONSOLE in RStudio, press Enter/Return, and wait several minutes until you get a “>” prompt at the bottom of the R Console):
if (!require('manipulate')) utils::install.packages('manipulate')
if (!require('remotes')) install.packages('remotes'); require('remotes')
remotes::install_github('droglenc/NCStats',type='binary')You can test whether the NCStats package installed correctly by running the line below (again, copy the line, paste it into the CONSOLE in RStudio, and press Enter/Return).
library(NCStats)If the installation is successful you should get about 20 lines of results in the console. As long as none of those lines says “Error” then the install should be correct. You can further test the install by running the first line below to see if you get the results on the two lines below that.
Summarize(1:3)
n mean sd min Q1 median Q3 max
3.0 2.0 1.0 1.0 1.5 2.0 2.5 3.0If your installation failed then please contact me ASAP for help.
RStudio-in-the-Cloud
Analyses with R and RStudio can be run through a web browser on “the cloud”, thus eliminating the need to install these softwares and the associated packages on your computer. Follow the directions below if you would like to use R/RStudio in the cloud. Only follow these directions if you don’t plan to use R/RStudio on your own computer and did not install R/Studio/NCStats as described in the three sections above.
You must create an RStudio Cloud account to use this service. Go to the RStudio Cloud website and select “Get Started.” In the ensuing dialog box enter your e-mail adddress, a password, and your first and last names. Press “Sign up.” In a subsequent screen you will be asked to provide a name for the account with prompts to use your first and last name. Just press “Create Account” here.
Note that the free version of RStudio Cloud is limited to 25 hours of work per month. In past semesters this has been adequate for our work up to the modules under the “Using R” subsection (see modules list). Once we start the “Using R” portion of the modules you will likely need to upgrade to the “Cloud Plus” account which is currently $5/month for 75 hours of project time. Thus, the free account should be adequate for about the first 2/3rds or so of the semester, so don’t sign up for the “Plus” account until later in term.
Once you have created your account, follow this link to the “MTH107” project, which should open a project that looks close to that below. Currently this project is in my “workspace.” Click on the red “Save a Permanent Copy” item in the top bar to save the project to your workspace (see below). When you return to “RStudio in the Cloud” in the future this project will appear in the left-hand menu, open it to begin your work.
R Basics
Saving
Results are not saved in R. Rather, “scripts” of successful R commands are saved and, then, if the analysis needs to be re-done, the entire set of commands is opened and run again. When writing a report, all tabular and graphical output should be copied from R and pasted into your report document. This document will serve as your analysis report and can be modified to include answers to questions, references to the tables and graphs, etc. Most data used in this class will be entered into R through “comma-separated values” text file (discussed later).
R does allow one to save a “workspace”, though I urge you not to do that. Rather, save your “good” commands in a script and save your “good” results in a report document; do not save the workspace.
Expressions and Assignments
Expressions in R are mathematical “equations” that are evaluated by R with a result seen immediately. An example of an expression in R is
5+log(7)-pi
#R> [1] 3.804317where log() and pi are built-in functions used to compute the natural log and find the value of \(\pi\), respectively. Expressions in R are like using a calculator where the result is shown, but not saved for subsequent analyses. In addition, expressions in R follow the same order of operations and use of parentheses as expressions entered into your calculator.
Results from an expression are typically saved for further computations by assigning the results to an object with the assignment operator (i.e., <-). The general form for saving the result of an expression into an object is object <- expression. The result of the expression will not be seen unless the object name is subsequently typed into R (but see below). For example, the result of the previous expression is saved into an object called x and then viewed with
x <- 5+log(7)-pi
x
#R> [1] 3.804317The result of an expression can be both assigned and printed by surrounding the command in parentheses. For example, the following assigns the result of the expression to y and prints the result.1
( y <- 15*exp(2) )
#R> [1] 110.8358An object can be named whatever you want, with the exception that it cannot start with a number or symbol (e.g., !, @), contain a space, or be the name of a reserved word or function (e.g., pi or log). Object names should be short and simple enough that you can remember what is contained in the object. It is also good practice to view the object immediately after making the assignment to make sure that it contains results that seem appropriate.
Functions and Arguments
R contains many “programs,” or functions, to perform particular tasks. A function is “called” by typing the function name followed by open and closed parentheses.2 The log() function, used in the previous section, is an example of a function. Arguments, which the function will use to perform its task, are contained within the parentheses. In log() above, the argument is the number in parentheses (and for which the natural log will be computed). Many other functions will be described below and in subsequent modules.
NCStats Package
NCStats is a set of R functions that makes the statistical methods used in this class easier. You need to load the NCStats package at the beginning of every R script used in this class. Packages are loaded by including the package name within library().
Corrections
If you see any errors (code, typographical, or logical) please bring these to my attention.