Is there a difference between assigning value using '<-' or '=' in R?

Question

In practical terms there seems to be no difference, but underneath the cloth as language processes there is some difference?

Answer 1

There is no difference in the vast majority of cases. Commands

x <- 5

and

x = 5

are identical.

However, if you want to assign arguments to a function, you are required to use =. For example, to generate a sample of 10 observations of a normal random variable with mean 5 and standard deviation 2, only the command

rnorm(10, mean = 5, sd = 2)

works. No use trying to rotate

rnorm(10, mean <- 5, sd <- 2)

that you will not get the desired result.

---

Particularly, I prefer to use <- whenever I’m going to assign some value to an object. I find it more elegant, because it differentiates from assigning arguments to functions. But it’s all about style anyway.

Answer 2

Complementing Marcus' answer. An interesting point is precedence of those operators. The <- comes before =. What makes that:

> a <- b = 1
Error in a <- b = 1 : não foi possível encontrar a função "<-<-"

It doesn’t work, but:

> a = b <- 1
> a
[1] 1
> b
[1] 1

Work.

This is the only case I’ve ever seen wrong... But as it is very unusual, it turns out whatever is used.

In fact, I just saw that this example is in pecedência.

## '=' has lower precedence than '<-' ... so you should not mix them
##     (and '<-' is considered better style anyway):
## Consequently, this gives a ("non-catchable") error
 x <- y = 5  #->  Error in (x <- y) = 5 : ....

If you call code style, most books recommend using <-. Example:

Advance R

Assignment Use <-, not =, for assignment.

# Good
x <- 5
# Bad
x = 5

Google R Style Guide

Assignment

Use <-, not =, for assignment.

GOOD:

x <- 5

BAD:

x = 5

---

Bonus

In Rstudio, you can use Alt + - to do the <-.