No, it is not. In order for a relation to be a function, each input can have only one output. Yet in order for the inverse relation to be a function, each output (of the original function) can only have 1 input that maps to it. Yet this clearly isn't the case for the function because the output 1/2, for example, has more than one input (angle) (ex's: and ) that map to it. (In fact, there exist infinitely many angles whose cosine ratio is 1/2.) Therefore, the inverse relation is not a function.