Numbers, text and logical values

R distinguishes what kind of value a vector holds, because that decides what you can do with it. You can calculate with numbers, not with text. Three types are enough to begin with.

Text always goes in quotation marks. TRUE and FALSE go without, because they are not text but logical values.

Asking for the type

class() tells you which type a value has:

diet <- c("vegan", "omnivore", "vegetarian")
class(diet)

Whole numbers that R produces itself, for example with 1:10, have the type integer. For you they behave like any other number.

One vector, one type

A vector can have only one type. If you mix numbers and text, R silently turns everything into text, because a number can always be written as text, but not every text as a number:

answers <- c(3, 5, "no answer", 2)
answers

The quotation marks in the output show it: the numbers have become text, which you can no longer calculate with. This happens often when reading data, when a single word sits in a column of numbers.

R does the same with numbers and logical values:

values <- c(5.4, TRUE, FALSE)
values

TRUE has become 1 and FALSE 0. R always looks for the lowest common denominator, the type into which all values can be converted. A logical value can be read as 0 or 1, but what would 5.4 be as a logical value? That is why the order is fixed: logical values become numbers and numbers become text, never the other way round.

Converting

With as.numeric() you turn text back into numbers. Whatever cannot be read as a number becomes NA, a missing value, and R warns you about it:

as.numeric(c("3", "5", "no answer", "2"))

There is also as.character(). The other way round, R counts TRUE as 1 and FALSE as 0 when you calculate with logical values; you already used that for counting with comparisons.

When R is to calculate with text

If you try to calculate with text, R stops and says that an argument is not numeric:

"3" + 4

When you meet this message, check with class() which type your values really have.

Key points

• Three types to begin with: numbers (numeric), text (character) and logical values (logical); class() tells you the type.
• A vector has only one type; mixed values become text.
• as.numeric() converts and turns anything unreadable into NA.