5  Reading data

5.1 Learning Objectives

This section covers different methods for importing and exporting data in R, by the end of this chapter you will be able to:

  • Import CSV and Excel files using base R, readr, and readxl.

  • Explain the difference between data frames and tibbles.

  • Use the here package for reproducible file paths (no setwd()!).

  • Paste small tables using datapasta.

  • Save and reload R-specific file formats (.RDS, .RData).

  • Use purrr::map() to read many data files automatically.

5.2 readr

If you’ve used R before, you might wonder why we’re not using read.csv(). There are a few good reasons to favour readr functions over the base equivalents:

  • They are typically much faster (~10x) than their base equivalents. Long running jobs have a progress bar, so you can see what’s happening. If you’re looking for raw speed, try data.table::fread(). It doesn’t fit quite so well into the tidyverse, but it can be quite a bit faster.

  • They produce tibbles, they don’t convert character vectors to factors, use row names, or munge the column names. These are common sources of frustration with the base R functions.

Feature Base data.frame Tidyverse tibble
Printing Prints all rows Prints compactly
Strings Become factors Stay as characters
Subsetting May simplify results Keeps consistent structure
Created by read.csv() read_csv()

5.2.1 useful options

readr::read_csv("files/penguins_raw.csv",
         col_types = cols(`Body Mass (g)` = col_double(),
                          `Sex` = readr::col_factor(levels = c("MALE", "FEMALE")),
                          `Date Egg` = readr::col_date("%d/%m/%Y")))

# Fixing character strings as factors
# Fixing dates at import requires you to specify date orders and separators

5.3 Importing from excel

library(readxl)
penguins_rawxl <- read_excel("data/raw/penguins_raw.xlsx", sheet = "Sheet1")

5.3.1 useful options

Argument Use Example
sheet Select by name or number sheet = 2
range Import only part of sheet range = "A1:D20"
skip Ignore header rows skip = 2
col_names Set your own names col_names = c("id", "age", "group")
col_types Set column types manually col_types = c("text", "numeric", "guess")
na Define missing values na = c("", "NA", "missing")

5.4 Filepaths

To maintain a clean and efficient workflow in R, it’s advisable to avoid using setwd() at the beginning of each script. This practice promotes the use of safe file paths and is particularly important for projects with multiple collaborators or when working across different computers.

Important

Why we use projects

5.4.1 Absolute vs. Relative Paths:

While absolute file paths provide an explicit way to locate resources, they have significant drawbacks, such as incompatibility and reduced reproducibility. Relative file paths, on the other hand, are relative to the current working directory, making them shorter, more portable, and more reproducible.

An Absolute file path is a path that contains the entire path to a file or directory starting from your Home directory and ending at the file or directory you wish to access e.g.

C:/home/your-username/project/data/penguins_raw.csv

  • If you share files, another user won’t have the same directory structure as you, so they will need to recreate the file paths

  • If you alter your directory structure, you’ll need to rewrite the paths

  • An absolute file path will likely be longer than a relative path, more of the backslashes will need to be edited, so there is more scope for error.

A Relative filepath is the path that is relative to the working directory location on your computer.

When you use RStudio Projects, wherever the .Rproj file is located is set to the working directory. This means that if the .Rproj file is located in your project folder then the relative path to your data is:

data/penguins_raw.csv

This filepath is shorter and it means you could share your project with someone else and the script would run without any editing.

5.4.2 The here Package:

To further enhance this organization and ensure that file paths are independent of specific working directories, the here package comes into play. The here::here() function provided by this package Müller (2025) builds file paths relative to the top-level directory of your project.

In the above project example you have raw data files in the data/raw directory, scripts in the scripts directory, and you want to save processed data in the data/processed directory.

To access this data using a relative filepath we need:

raw_data <- read.csv("data/raw/penguins_raw.csv")

To access this data with here we provide the directories and desired file, and here() builds the required filepath starting at the top level of our project each time

library(here)

raw_data <- read.csv(here("data", "raw", "penguins.csv"))

5.5 datapasta

The datapasta package lets you paste small tables directly into R — perfect for quick tests or examples.

library(datapasta)
# Copy a small table (e.g. from Excel), then run:
dpasta()

It automatically pastes code like this:

tribble(
  ~id, ~age, ~group,
  1, 23, "control",
  2, 27, "treatment"
)

5.6 R data types

.RDS files, or R Data Serialization file, is a binary file format in R used to save individual R objects.

# Create some sample data
my_data <- data.frame(
  ID = 1:3,
  Name = c("Alice", "Bob", "Charlie"),
  Score = c(95, 87, 92)
)

# Save the data frame to an .RDS file
saveRDS(my_data, file = "data/clean/my_data.RDS")

# Clear the current workspace
rm(list = ls())

# Load the data frame from the .RDS file
loaded_data <- readRDS("data/clean/my_data.RDS")

# Access the loaded data
print(loaded_data)

Your turn

5.6.1 Question

When might saving data to RDS file format be useful?

5.7 Reading multiple files

Important

This section uses several concepts we haven’t really been introduced to yet. Including writing functions, iterative programming and string matching.

Here we actually start with a complete dataframe - and first iterate to split into 25 equally sized dataframes.

walk2 operates in the same way as map2 - but is the preferred option here as it is “silent”

fs::dir_create("data/many_files")

unique_species <- penguins_raw |> 
  distinct(Species) |> 
  pull()

peng_samples <- map(unique_species, 
                    ~ filter(penguins_raw, `Species`==.x)
                    )

map2(peng_samples, unique_species, 
     ~ write_csv(.x, paste0(glue::glue("data/many_files/{.y}.csv")))
     )

5.7.1 Create a vector of file paths

Now, to create a vector of file paths, we’ll use the list.files function in R. This function allows us to identify and list all the files with a specific extension in a directory. In this example, we’re looking for CSV files in the “data/many_files” directory.

list_files <- dir_ls(path = "data/many_files",
                                    pattern = "csv", full.names = TRUE)
data/many_files/Adelie Penguin (Pygoscelis adeliae).csv
data/many_files/Chinstrap penguin (Pygoscelis antarctica).csv
data/many_files/Gentoo penguin (Pygoscelis papua).csv

5.7.2 Read multiple files

Now that we have obtained the file paths, we can proceed to load the files into R. The preferred method in the tidyverse is to use the map_dfr function from the purrr package. This function iterates through all the file paths and combines the data frames into a single, unified data frame. In the following code, .x represents the file name or path. To read and output the actual content of the CSV files (not just the filenames), you should include .x (the path) within a readr function. While this example deals with CSV files, this approach works similarly for other rectangular file formats.

penguins_data_test <- map_dfr(list_files,
              ~ read_csv(.x))

5.7.3 Selecting files

Now, to filter and choose specific files for reading, we’ll use the str_detect() function from the stringr package in R. This function allows us to search for specific patterns within our vector of file paths and select files that match our criteria.

The negate argument ensures that we only select files that either do or don’t match the pattern. This work is made easier when we have good naming conventions.

list_files[str_detect(list_files, pattern = "Gentoo",
negate = TRUE)]
data/many_files/Adelie Penguin (Pygoscelis adeliae).csv
data/many_files/Chinstrap penguin (Pygoscelis antarctica).csv"

5.8 Summary

  • Import data using readr and readxl.

  • Understand tibbles vs data frames.

  • Keep file paths reproducible with here().

  • Paste, save, and reload data flexibly.

  • Automate imports with purrr::map().