rmodel.matrix

How to remove unused levels from model matrix

I created a model matrix. Some variables are categorial variables. After filtering the data some levels are not in the data set anymore. How can i remove the unused levels? Can i apply the factor function on the categorial variables ?

Solution

  • you can use the droplevels function in base R. Let x be your factor/ categorical variable:

    x <- as.factor(c("cat", "dog","cat", "gator"))
x
# [1] cat   dog   cat   gator
# Levels: cat dog gator

# somewhere in analysis you removed the only entry for a level
x <- x[x!= "gator"]     
x
# [1] cat dog cat
# Levels: cat dog gator

droplevels(x)
# [1] cat dog cat
# Levels: cat dog

    Refer droplevels R documentation for more details.