Why a quadratic needs its domain cut to have an inverse — quick review
Picture a machine that squares numbers.
A quadratic f(x) = a(x − h + k, with a ≠ 0, has its vertex at (h, k). On all real numbers it is not a one-to-one function, so it has no inverse function. On the restricted domain x ≥ h (the right half) or x ≤ h (the left half) it is a one-to-one function, so it has an inverse function. The range does not change: y ≥ k when a > 0, and y ≤ k when a < 0.
Find the vertex (h, k). In the form a(x − h + k, read it off, remembering that the sign inside the parentheses is the opposite of h. In the form a + bx + c, compute h = −, then k = f(h).
Picture the vertical line x = h through the vertex. It splits the parabola into a left half and a right half that mirror each other.
Choose one half. If the problem names a side, use that side. If it does not, keep the right half.
Write the restricted domain: x ≥ h, which is [h, ∞), for the right half, or x ≤ h, which is (−∞, h], for the left half. The vertex stays in either way.
Write the range. If a is positive the parabola opens up and the range is y ≥ k. If a is negative it opens down and the range is y ≤ k.
Check with two inputs the same distance from h, such as h − 1 and h + 1. They give the same output, and only one of them is still allowed.
Find the vertex (h, k).
Pick the side to keep: the one the problem names, or the right side if it names none.
Write the restricted domain: x ≥ h or x ≤ h, with the vertex included.
Write the range from k: upward if the parabola opens up, downward if it opens down.
Test two inputs that mirror each other across x = h. Only one of them may still be allowed.
Let f(x) = (x − 2 + 1. First show that f, used on all real numbers, is not a one-to-one function. Then cut its domain at the vertex, keeping the right half, and give the restricted domain and the range. In plain words: find two different inputs with the same output, then choose which inputs to keep so that this can never happen.
On all real numbers f is not a one-to-one function, because f(0) = 5 and f(4) = 5.
Restricted domain: x ≥ 2, or [2, ∞).
Range: y ≥ 1, or [1, ∞).
On this restricted domain f is a one-to-one function, so it has an inverse function.
Cutting at x = 0 out of habit. That works only when the vertex sits on the y-axis, as in or + 5. For (x − 4, the inputs 3 and 5 are both positive and both give 1, so x ≥ 0 fixes nothing. Find the vertex first, every time, and cut at its x value h.
inverse function
The function that undoes another function. If f turns 3 into 7, then the inverse, written , turns 7 back into 3.
one-to-one function
A function in which every output comes from only one input. No two inputs share an output, so no horizontal line meets its graph twice.
restricted domain
A domain made smaller on purpose. You keep only part of the inputs, such as x ≥ 0, so that the function becomes one-to-one and can have an inverse.