Reading data into R

Andrew Irwin, a.irwin@dal.ca

2026-02-03

Reading data

Most data originates outside of R
To use data with R, you must
- Obtain a copy of the data in some format
- Get R to “read” those data
- Check that the data were interpreted correctly

What formats do data come in?

Spreadsheet (Excel, Google Sheets, Numbers, LibreOffice Calc)
Text files (csv, tab separated, delimited)
Binary formats (various)
Self-documenting (NetCDF for geophysical data, specialized for astronomy, oceanography, computer model output)
An almost infinite variety of custom formats

Examples

Create your own excel worksheet
Environment Canada weather: https://climate.weather.gc.ca/climate_data/daily_data_e.html?StationID=50620
CO2: https://www.esrl.noaa.gov/gmd/webdata/ccgg/trends/co2/co2_mm_mlo.txt

Spreadsheet

library(readxl)
dt1 <- read_excel("static/L11/test-data.xlsx", sheet=1)
dt1 |> kable()

Name	Birthdate	Mass_kg	Species	Friendly
A nonymouse	2001-01-01	14.5	NA	FALSE
Frank	2008-09-01	4.1	Cat	TRUE
Boojum	1982-07-11	7.2	Dog	TRUE

Comma separated values

library(tidyverse)
dt2 <- read_csv("static/L11/en_climate_daily_NS_8202251_2020_P1D.csv")
dt2 |> paged_table()

Tab separated values

With a header

dt3 <- read_table("static/L11/co2_mm_mlo.txt", 
                  comment="#", col_names = FALSE)
dt3 |> paged_table()

Clean-up

names(dt3) <- c("year", "month", "decimal_year", "co2_monthly", 
                "co2_deseasoned", "n_days", "sdev_days", "uncertainty_mean")
dt3 <- dt3 |> 
  mutate(n_days = dplyr::na_if(n_days, -1), 
         sdev_days = dplyr::na_if(sdev_days, -9.99),
         uncertainty_mean = dplyr::na_if(uncertainty_mean, -0.99))

Another way to convert columns

library(hablar)
dt3 <- dt3 |> convert(int(year), 
                      int(month), 
                      num(decimal_year:uncertainty_mean) )

Warning, hablar has some functions with same names as functions in dplyr. For example, need to use dplyr::na_if to use the masked na_if function.

Rstudio data import helper

Flle > Import dataset > From text (readr) …

Cleaning variable names

Variable (column) names that contain spaces, brackets, arithmetic symbols, and other symbols
These variable names must be enclosed in backticks (e.g., `Max Temp (°C)`)
clean_names in the janitor package transforms the names to make them easier to type

Original names: Longitude (x), Date/Time, Max Temp (°C)

Cleaned names: longitude_x, date_time, max_temp_c

Things to watch out for

Multi-row headings, missing headings
Extra rows at the top of the file
Missing data coded in an “interesting” way
Multiple tabs (sheets) in a spreadsheet or workbook
Date formats
Numeric data interpreted as text

Summary

read_excel to read from spreadsheets
read_csv, read_tsv, and read_delim to read many text files
Try the Rstudio data import tool. Notice the R code generated for you.
Always check your data table in R
Lots of ways for errors to arise
Practice and ask for help if you have trouble