22 Re-shape

22.1 Wide data

This is the most common approach

each real-world entity is represented by one single row
its attributes are represented through different columns

City	Population	Area	Density
Leicester	329,839	73.3	4,500
Nottingham	321,500	74.6	4,412

22.2 Long data

This is probably a less common approach, but still necessary in many cases

each real-world entity is represented by multiple rows
- each one reporting only one of its attributes
one column indicates which attribute each row represent
another column is used to report the value

City	Attribute	Value
Leicester	Population	329,839
Leicester	Area	73.3
Leicester	Density	4,500
Nottingham	Population	321,500
Nottingham	Area	74.6
Nottingham	Density	4,412

22.3 tidyr

The tidyr (pronounced tidy-er) library is part of tidyverse and it provides functions to re-shape your data

city_info_wide <- data.frame(
  City = c("Leicester", "Nottingham"),
  Population = c(329839, 321500),
  Area = c(73.3, 74.6),
  Density = c(4500, 4412)
)

kable(city_info_wide)

City	Population	Area	Density
Leicester	329839	73.3	4500
Nottingham	321500	74.6	4412

22.4 tidyr::gather

Re-shape from wide to long format

city_info_long <- city_info_wide %>%
  gather(
    -City, # exclude city names from gathering
    key = "Attribute", # name for the new key column
    value = "Value" # name for the new value column
  )

City	Attribute	Value
Leicester	Population	329839.0
Nottingham	Population	321500.0
Leicester	Area	73.3
Nottingham	Area	74.6
Leicester	Density	4500.0
Nottingham	Density	4412.0

22.5 tidyr::spread

Rre-shape from long to wide format

city_info_back_to_wide <- city_info_long %>%
  spread(
    key = "Attribute", # specify key column
    value = "Value" # specify value column
  )

City	Area	Density	Population
Leicester	73.3	4500	329839
Nottingham	74.6	4412	321500