khuntzberry

Data Types

Overview

  • Numeric
    • Numbers with decimals
    • Example: c(1.2, 2.6, 3.4, 2.2, 5.7)
  • Character
    • Values made up of strings or letters
    • Example: c(“Alabama”, “New Jersey”, “Texas”)
  • Factor
    • String or numeric variables stored as groups
    • Stored as integers with value labels you set
  • Integer
    • Whole numbers
    • Example: c(1, 2, 3, 4, 5)
  • Logical
    • TRUE or FALSE
    • Example: c(TRUE, FALSE, FALSE, TRUE)

Checking and Changing Data Types

Checking Data Types

  • Use class() function to check a data type

Changing Data Types

  • To change or coerce an object to another data type, the format is as follows:
    • as.numeric() converts to numeric
    • as.integer() converts to integers/whole numbers
    • as.character() converts to character
    • as.factor() converts to factor

Changing Factor Variables to Numeric

  • There is a trick for changing variables from factor to numeric
  • Remember: Factor variables are stored as numbers
  • The stored number is based on the order or your values
  • If you use as.numeric() on a factor variable made up of numbers, the stored number will be returned (which will be meaningless to you!)
  • To solve this issue, change the variable to a character variable first, then convert it to a numeric variable
  • Use this method:
  • Do NOT use this method:

Deep Dive: Factor Variables

What Are Factor Variables?

  • Factor variables can be either strings or numeric
  • However, in the background, all values are grouped and stored as a number
  • Below you can see:
    • That city is stored as a character variable with string values
    • State but is stored as a number
      • The string values are labels

Creating Factor Variables

  • You can change other data types to factor variables
  • Levels and labels can be customized
  • In the data frame shown below, states is a character variables:

Factor Conversion: “As Is” Method

  • As explained above, the as.factor() method can be used to change the states variable to a factor variables as is

Factor Conversion: New Labels

  • The state variable can also be changed to a factor variables with different labels
    • Example: Changing the state abbreviations to the full state names
  • To do this follow the structure below:
  • First argument: Variable you want to convert to a factor
  • Second argument: The original variable values
  • Third argument: The new labels you want to apply to the original values
    • Order matters! Make sure the levels and labels values are in the same order!

Factor Conversion: Ordered

  • Character variables can also be changed to ordered factors
  • This should be done when factor levels have are ordered
    • Example: The temp.group variable is ordered, High > Medium > Low
  • To do this follow the structure below:
  • Add ordered = TRUE argument to the factor() arguments