R - Factor (Category, Enumerated Type)

1 - About

The factor function is used to encode a vector as a factor (ie categorical data).

When used with a numeric or a date, a binning function will return a factor.

From numeric to a category (For instance, for an id)

A factor is also known as:

Factors can be unordered or ordered.

A factor is an integer vector where each integer has a label use the str function to see it.

Factors are treated specially by modelling functions like lm() and glm()

3 - Syntax

factor(
         v = character(), 
         levels, 
         labels = levels,
         exclude = NA, 
         ordered = is.ordered(x), 
         nmax = NA
      )

where:

  • v is a vector
  • levels is an optional vector containing the data domain where the order of the levels can be set. This is important in linear modelling because the first level is used as the baseline level.
  • labels is an optional vector of labels for the levels
  • exclude is a vector of values to be excluded when forming the set of levels
  • ordered is a logical flag to determine if the levels should be regarded as ordered.
  • nmax is an upper bound on the number of levels

4 - Management

4.1 - Simple Initialization

A factor of colours with 4 values and 3 levels

> x=factor(c("Green","Blue","Red","Green"))
> x
[1] Green Blue  Red   Green
Levels: Blue Green Red

We can see that a factor is a composition of labels and integer vector (2 1 3 2):

str(x)
 Factor w/ 3 levels "Blue","Green",..: 2 1 3 2
> unclass(x)
[1] 2 1 3 2
attr(,"levels")
[1] "Blue"  "Green" "Red" 

4.2 - Level

The same factor of colours as above but with only two colours in the level (domain). One value becomes NA. If you want NA as level see the how to section

> x=factor(c("Green","Blue","Red","Green"),levels=c("Green","Blue"))
> x
[1] Green Blue  <NA>  Green
Levels: Green Blue

You can get the levels with the levels function

levels(x)
[1] "Green" "Blue" 

4.3 - Label

A factor of colours with two colours levels and different level labels

> x=factor(c("Green","Blue","Green"),levels=c("Green","Blue"),labels=c("LabelGreen","LabelBlue"))
> x
[1] LabelGreen LabelBlue  LabelGreen
Levels: LabelGreen LabelBlue

4.4 - Exclude

A factor of colours with a colour excluded:

> x=factor(c("Green","Blue","Green"),exclude="Green")
> x
[1] <NA> Blue <NA>
Levels: Blue

5 - How to

5.1 - Count the number of element by level

with the table function:

> x=factor(c("Green","Blue","Red","Green"))
> table(x)
x
 Blue Green   Red 
    1     2     1 

5.2 - Have NA as level

If you want NA as a level (ie allow missing values)

> x = factor(c("Blue", NA), exclude = NULL)
> x
[1] Blue <NA>
Levels: Blue <NA>

5.3 - Transform it back as a vector

5.4 - Order

The default order is alphabetical.

  • The function reorder: Reorder Levels of a Factor

6 - Continuous to Factor

6.1 - Date to weekday

Example creation of a weekday factor

data_frame$CREATED_ON_WEEKDAY <- factor(weekdays(data_frame$CREATED_ON),levels=c("Monday","Tuesday","Wednesday","Thursday","Friday","Saturday","Sunday"))

6.2 - Number to bin

df$ageFactor <- cut(df$age, breaks=c(0, 15, 45, 56, Inf))

Since even the integers are converted to character strings, they are sorted in a dictionary order (rather than by magnitude).

7 - Documentation / Reference

lang/r/factor.txt · Last modified: 2017/11/16 22:59 by 162.158.106.4