Importing/exporting files in R

The first step with data processing is to load the data into R’s workspace. R provides a large set of options to do so, and extensions exist that allow R to read in data files from other programs, including Excel, SPSS, SAS, and Matlab.

Table of contents

1. Working with .RData files
2. Working with .txt files
3. Working with .csv files
4. Read in .xlsx files

🡻

1. Working with .RData files

R allows you to save the user-defined objects present in your current workspace as an ‘.RData’ file, and to easily re-load these files at a later point:

# Example user-defined objects
vec <- 1:3
string <- c( 'Cats', 'are', 'cute' )
df <- data.frame( A = vec, B = string, stringAsFactors = F )

# Save everything in workspace in current working directory
save.image( file = 'Filename.RData' )
# Save specified objects to a desired sub-folder
save( df, file = 'Sub_folder/Filename.RData' )

# Load in previous data
load( file = 'Filename.RData' )
load( file = 'Sub_folder/Filename.RData' )

Note: A useful approach is to write a script to read the raw data in from its original format, process and clean the data, and then save the processed data (e.g., the resulting data frame) as a .RData file. In this way, you preserve the original raw data, but have a cleaned set of data that is particularly easy to read in via R.

🡹 🡻

2. Working with .txt files

One can create simple text files using R via the ‘write’ function. A whole new text file can be created (or an old one overwritten), or lines can be added to an existing file. This provides a simple tool to, for example, log the results of a script for debugging purposes.

# Create a vector of text
string = c( 'Hello world!', 'Here is a new line', 'Ad infinitum' )
# Save a text file in the current working directory
# (each string in the character vector is added as a new line)
write( string, file = 'Example.txt' )
# Add another line of text to the existing file
write( 'Post-hoc line', file = 'Example.txt', append = T )

Text files can be read into R as a character vector via the ‘scan’ command:

# Create a text file
write( 'Hello world\nThis is a new line', file = 'Example.txt' )
# Read the text file back into R
chr_vec <- scan( file = 'Example.txt', what = 'character' )
# Each word read in as an element
print( chr_vec )
# Instead read in each line as an element
chr_vec_2 <- scan( file = 'Example.txt', what = 'character', sep = '\n' )

Note: It is important to indicate to R that you will be reading in character strings. By default, ‘scan’ expects only numbers, and will throw an error if it encounters otherwise.

There will cases in which you have data that is not formatted as a spreadsheet; ‘scan’ provides a flexible, robust way to still read in such data for subsequent processing.

However, it is easier to create and load data in spreadsheet form, via the commands ‘write.table’ and ‘read.table’:

# Create data frame
df <- data.frame(
  Animal = c( 'Cat', 'Dog', 'Monkfish' ),
  Ranking = 1:3,
  stringsAsFactors = F
)
# Save as a '.txt' spreadsheet
write.table(
  df, # Data frame or matrix
  file = 'Example.txt', # File name
  sep = ' ', # Delimiter separating columns
  row.names = F, # Don't include row labels
  quote = F # Don't place quotations around entries
)
# Read text file back in as a data frame
df_2 <- read.table(
  file = 'Example.txt', # File name
  sep = ' ', # Delimiter separating columns
  header = T, # First line used for column names
  stringsAsFactors = F
)

The ‘write.table’ and ‘read.table’ commands provide general-purpose functions to read in spreadsheets, and allow the user a high degree of control over aspects such as the delimiter separating columns. R has even more specialized functions for common file types, in particular comma-delimited files.

🡹 🡻

3. Working with .csv files

Comma-delimited files (.csv) provide a very easy data format to read in or create, and are especially useful when you need to save data in a format that can be processed by a wide variety of programs. R provides the ‘write.csv’ and ‘read.csv’ functions specifically for these type of files. These functions work similarly to ‘write.table’ and ‘read.table’, but assume the delimiter is a comma:

# Create data frame
df <- data.frame(
  Food = c( 'Cake', 'Ice cream', 'Vegetables' ),
  Ranking = 1:3,
  stringsAsFactors = F
)
# Save as a '.csv' file
# No longer necessary to indicator column delimiter
write.csv(
  df, 
  file = 'Example.csv', 
  row.names = F, # Don't include row labels
  quote = F # Don't place quotations around entries
)
# Read text file back in as a data frame
df_2 <- read.csv(
  file = 'Example.csv', 
  header = T, # First line used for column names
  stringsAsFactors = F
)

🡹 🡻

4. Read in .xlsx files

Excel spreadsheets (i.e., .xls and .xlsx) are a proprietary format, and R can’t by default read in these types of files. Fortunately, several R packages exist that resolve this issue. For example, Hadley Wickham’s ‘readxl’ package provides straightforward tools to read in Excel files without requiring any external dependencies:

# Install package
install.packages("readxl")
# Load in package
library( readxl )

# When an Excel file file has multiple sheets, 
# indicate which sheet to read in via the 
# sheet name or by its index number

# Specify sheet by its name
df <- read_excel( 
  path = "Example.xlsx", 
  sheet = "Name",
  na = "" # Allows customizing of what to treat as missing data
)
  
# Specify sheet by its index
df <- read_excel( 
  path = "Example.xlsx", 
  sheet = 2
)

🡹

Return to: Data processing; Sections; Home page