Mod or rem. Which one to use with positive integers in Haskell?

user_4239355 0
Viewed 374

The rem in Haskell behaves just like the % in JavaScript. When using the clock analogy rem always goes clock-wise regardless if the value is positive or negative. rem takes the sign of the divident

divident / divisor = quotient

Prelude> 10 `rem` 3
1

Prelude> -10 `rem` 3
-1

% in JavaScript works the same way

$ node

> 10 % 3
  1

> -10 % 3
  -1

mod behaves like rem when both values are positive. The only difference is that when one of the numbers is negative mod goes backwards on the clock, not clock-wise and takes the sign of the divisor.

Knowing that mod and rem give the same result when both numbers are positive which one is recommended to be used in Haskell?

haskell
1 Answers

div and mod are more semantically sensible, while quot and rem are marginally faster. So my advice is to use div and mod as a reflex, and only ever change to using quot and rem after doing a lot of work to make sure their performance is actually relevant to some slowdown you absolutely must fix and doing a lot of work to prove that it's safe. In the long run, subtle sign bugs are almost certainly going to cost you much more debugging time than subtle performance bugs will cost you computation time.

user_7916040
Related