Confusing the inverse with the reciprocalf−1(x) is different from 1/f(x).
Forgetting to swap x and yif we do not swap, we get back the original function.
Ignoring the domainthe domain of the inverse is the range of the original function, not its domain.
Taking both roots in a restricted quadraticchoose only the root that fits the given domain.