Introduction to Data Analysis in R

This course is aimed at the intermediate R developer who wants to learn how to do useful data analysis tasks in R. It will focus on “modern R”, specifically using the tidyverse collection of packages which are designed for data science.

Data analysis is a huge topic and we couldn’t possibly cover it all in one short course. The purpose of this workshop is to give you an introduction to some of the most useful tools and to demonstrate some of the most common problems that surface.

This workshop assumes you have knowledge of R equivalent to that covered in Beginning R and Intermediate R. This workshop provides the background necessary to follow Text and Sentiment Analysis in R.

You can jump ahead to any chapter:

walkthrough video for this page

For the purposes of this workshop we will be using RStudio. If you haven’t installed and got RStudio working, then please follow the instructions in the Intermediate R workshop.

In previous courses, you’ve used the Rscript command line program to execute scripts and the RStudio console to run interactively. This course will use another tool called R Notebooks (or R Markdown Notebooks) to run your R code. It operates like a standard RStudio console interactive session, with the addition of allowing you to intersperse your code with blocks of text to explain what you’re doing and embed output such as graphs directly into the page.

To get started, launch RStudio. In the menu, select “File | New File | R Notebook”. This will open a new R Notebook, which should look somethign like this;

The notebook is initially untitled. Give it a name by saving it (either CTRL or CMD + S, or click on the “Save” icon). Call it something like hello.Rmd (Rmd is the extension used for R Markdown Notebooks).

Throughout this course you will likely want to start a new notebook for each section of the course so name them appropriately to make it easier to find them later.

Getting started

Once the notebook is launched, you will see a lot of example text. Clear everything except for the title at the top, e.g. as in

The notebook is just a text file, which is formatted using markdown. This is a simple markup language that lets you write rich documents.

You can add R code to this file by placing into R code blocks. You can do this by clicking “Insert | R”, e.g. as in;

When you click this, it will add the following text to your R notebook;

(note, you can type this yourself if you want - you don’t need to click “Insert | R”)

You can write any R code you want into this block. For example, type;

cat("Hello R!\n")

Next, click on the “Run current chunk” button (the green triangle on the right)

This will run the R code in the R block, printing the output directly on the screen.

Pretty cool!

You can also write text, for example now write;

This is some text

## And this is a heading

Next, add another R block, e.g.

a <- 5
b <- 7
a + b

(remembering to use “Insert R” or typing in the ```{r})

and then click “Run”. You should now see;

The R blocks in a notebook are linked together so a variable defined in one is available in all the R blocks from that point on. So in the second cell you can use the variables a and b:

a - b

Most R functions produce outputs that will look really nice in an R notebook, for example a tibble will format nicely, while a graph (which we will introduce later) will render in place, e.g.

library(tidyverse)
tibble(cars)

will give a nice interface to a tibble of the built-in cars dataset, while,

plot(cars)

will plot the graph of that dataset.

Finally, this document can be rendered to a nice web page or a good-looking PDF or Word file. You can do this by clicking the “Preview” button.

It is becomming common for people to use R notebooks as a sort of lab notebook where they document their processes, interspersed with code. This style of working where you give prose and code equal weight is sometimes called literate programming.

EXERCISE

Take the following code and break it down, chunk by chunk, interspersing it with documentation explaining what each part does using text:

prices <- c("apple"=0.40, "banana"=0.50)
my_purchase <- c(
    "apple"=1,
    "banana"=6
)
grocery_bill <- 0
for (fruit in names(my_purchase)){
    grocery_bill <- grocery_bill + prices[fruit]*my_purchase[fruit]
}
cat(sprintf("I owe the grocer %.2f\n", grocery_bill))

You don’t need to put only one line of code per cell, it makes sense to group some lines together.

Throughout this course, use the R notebook to solve the problems. Follow along with the examples, typing them into your own notebooks and see how they work.

Introduction to Data Analysis in R

Getting started

Further reading

Next