Getting started with data in R

Eirini Zormpa

The RSA

Last time you learned how to:

  • ✅ Interact with the RStudio GUI,
  • ✅ Set up projects,
  • ✅ Create files from RStudio,
  • ✅ Assign values to objects,
  • ✅ Use functions,
  • ✅ Create and subset vectors,
  • ✅ Work with missing data

Questions from last time?

Learning objectives

  • Read data into R.
  • Understand and manipulate data.frames or tibbles.
  • Understand and manipulate factors.
  • Alternate between date formats.

Data frames

data.frames are the standard data structure for tabular data in R.

They look very similar to spreadsheets (like in Excel) but each column is, in fact, a vector:

  • Each vector needs to be of the same length, for a perfect rectangle.
  • Because the columns are all vectors, they must all be of the same type.

A note on terminology

Technically, what we will be working with in these workshops aren’t data.frames, they are tibbles. tibbles are basically dataframes for the tidyverse - they have some subtle differences but nothing you need to worry about at this point.

Tabular data: What is tidy data?

Tabular data: Why tidy data?

Tabular data: File formats

Comma-delimited (.csv)

  • 👍 commonly used
  • 👎 annoying when data itself contains commas

Tab-delimited (.tsv)

  • 👍 no confusion when data contains commas or semicolons
  • 👎 not very commonly used (at least not yet)

The data

⚠️ NOT REAL DATA ⚠️


The data have been modified from another dataset to mimic ONS Census data. Their sole purpose is to be used in training.

The data: variables

variable description
ID a number to identify the participant
region where in the UK the participant is located
interview_date the date the interview took place
household_size the number of members in the household
age the ages of the people in the household
dwelling_type the type of dwelling
bedrooms the number of bedrooms in the dwelling
central_heating whether the dwelling has central heating
cars the number of cars the participant owns
community_establishment the types of community establishment in the area
religion the participant’s religion

Importing data: Folders

  1. Double click on the R Project you created for the workshop to open RStudio.
  2. Check that the files you see in your Files tab are the right ones (you should only see the scripts folder and the .Rproj file)
  3. Go to the console and type the following commands
# create separate folders for the raw and clean data
dir.create("data_raw")
dir.create("data_clean")

# only if you don't have one already, create a folder for the scripts
dir.create("scripts")

Importing data: Download

Then we need to 1) download the data and 2) save it in the data_raw folder we just created it.

We can do both in one go in R by typing the following command in the console:

# download the data
download.file(url = "https://raw.githubusercontent.com/theRSAorg/r-training/main/data_raw/synthetic-census-data.csv?download=1",
              destfile = here("raw_data", "census_data.csv"))

After you have run this command, open the data_raw folder and check that there is a file called census_data.csv.

Exercise 2.1

⏰ 10 mins

10:00
  • Create a new tibble (census_200) with the data in row 200 of census_data
  • Create a new tibble (census_last) from the last row, without typing out the row number
    • Check the output against tail()
  • Create a new tibble (census_middle) from the middle row of the dataset
  • Use the - notation to reproduce the behavior of head(census_data) (show rows 1-6)

Exercise 2.1 solution

## 1.
census_200 <- census_data[200, ]
## 2.
# Saving `n_rows` to improve readability and reduce duplication
n_rows <- nrow(census_data)
census_last <- census_data[n_rows, ]
## 3.
census_middle <- census_data[n_rows / 2, ]
## 4.
census_head <- census_data[-(7:n_rows), ]

Factors

R has a special data class, called factor, to deal with categorical data. Factors:

  • are stored as integers associated with labels, though they look like character vectors
  • can be ordered (ordinal) or unordered (nominal)
  • create a structured relation between the different levels (values) of a categorical variable, such as days of the week or responses to a question in a survey

Exercise 2.2

⏰ 5 mins

05:00
  1. Change the columns region and dwelling_type in the census_data data frame into a factor.
  2. Using the functions you learned before, can you find out…
  • How many different dwelling types there are in the dwelling_type column?
  • How many participants there are who are based in London?

Exercise 2.2 solution

census_data$dwelling_type <- factor(census_data$dwelling_type)
census_data$region <- factor(census_data$region)

nlevels(census_data$dwelling_type)
summary(census_data)

Exercise 2.3

⏰ 5 mins

05:00
  1. Rename “no”, “yes”, and “unknown” to “No”, “Yes” and “Unknown” respectively.
  2. Recreate the barplot such that “Unknown” is last.

Exercise 2.3 solution

census_data$central_heating <- fct_recode(census_data$central_heating,
                              No = "no",
                              Unknown = "unknown",
                              Yes = "yes")

census_data$central_heating <- factor(census_data$central_heating, levels = c("No", "Yes", "Unknown"))

plot(census_data$central_heating)

Dates

To avoid ambiguity, use the RFC3339 standard: YYYYMMDD (or YYYY-MM-DD).

Summary

  • ✅ Read data into R.
  • ✅ Understand and manipulate tibbles.
  • ✅ Understand and manipulate factors.
  • ✅ Alternate between date formats.