Description
Math functions have both a "type" and a "return type" defined, which are subtly different. It's not clear to me whether this is an intentional distinction or not. I started writing this issue because I thought it was a bug but now I've stared at it for a while I'm not sure. Thought it was better to raise the issue anyway.
As an example:
The math function type summary says:
sign(), sin(), cos(), tan(), pow(), sqrt(), log(), exp()
«[ ]» (empty map).
But the prose defining sin() says:
The sin(A), cos(A), and tan(A) functions [...] They all represent a
<number>, with the return type made consistent with the input calculation’s type.
If I'm reading that right, it means the return type is either «[ ]», or «[ ]» with a percent hint.
Some clarification would be helpful. Thanks!