Quarry School

College Algebra / Inverses and radical functions

Inverses and radical functions

Subject: Mathematics

All 6 lessons, in order.

The unit at a glance

  1. Why a quadratic needs its domain cut to have an inverse

    Picture a machine that squares numbers.

  2. Finding the inverse of a quadratic on a cut domain

    Think about putting on socks, then shoes.

  3. Inverses of other polynomials

    A polynomial is a sum of terms such as 2x3 or −5, where x carries whole-number powers.

  4. Radical functions and their domains

    A radical function is a rule with x under a root sign, such as 2x+6 or x3.

  5. Inverses of rational functions

    A rational function is a fraction whose top and bottom are polynomials.

  6. A function and its inverse undo each other

    Think of a coat check.

Rules (6)
  • Why a quadratic needs its domain cut to have an inverse

    A quadratic f(x) = a(x − h)2 + 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.

  • Finding the inverse of a quadratic on a cut domain

    For f(x) = a(x − h)2 + k on the restricted domain x ≥ h, the inverse function is f⁻¹(x) = h + (x−k)÷a. On the restricted domain x ≤ h it is f⁻¹(x) = h − (x−k)÷a. The domain of f⁻¹ is the range of f, and the range of f⁻¹ is the restricted domain of f.

  • Inverses of other polynomials

    A cubic of the form f(x) = a(x − h)3 + k, with a ≠ 0, is a one-to-one function on all real numbers, so it has an inverse function with no restricted domain. To find it: write y = f(x), swap x and y, and solve for y, using a cube root to undo the cube. A cube root has one value and takes no ± sign. The result is f−1(x) = ∛(x−ka) + h, with domain (−∞, ∞) and range (−∞, ∞).

  • Radical functions and their domains

    Square root (and every even root): f(x) = ax+b gives a real number only when ax + b ≥ 0. Solve that inequality. Its solution is the domain.
    Cube root (and every odd root): g(x) = ∛(ax + b) gives a real number for every x. Its domain is (−∞, ∞).

  • Inverses of rational functions

    For f(x) = ax+bcx+d with c ≠ 0 and ad − bc ≠ 0, swap and solve gives f⁻¹(x) = −dx+bcx−a. Domain of f: x ≠ −dc. Range of f: y ≠ ac. Domain of f⁻¹ = range of f. Range of f⁻¹ = domain of f.

  • A function and its inverse undo each other

    Two functions f and g are inverses of each other exactly when g(f(x)) = x for every x in the domain of f, and f(g(x)) = x for every x in the domain of g. Then g = f⁻¹, the domain of g is the range of f, the range of g is the domain of f, and the point (a, b) is on the graph of f exactly when (b, a) is on the graph of g. The two graphs are mirror images across the line y = x.

Recipes (6)
  • Why a quadratic needs its domain cut to have an inverse
    1. Find the vertex (h, k). In the form a(x − h)2 + k, read it off, remembering that the sign inside the parentheses is the opposite of h. In the form ax2 + bx + c, compute h = −b2a, then k = f(h).
    2. 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.
    3. Choose one half. If the problem names a side, use that side. If it does not, keep the right half.
    4. 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.
    5. 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.
    6. 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.
  • Finding the inverse of a quadratic on a cut domain
    1. Write y in place of f(x), and write the restricted domain (the cut domain) beside it as a note, for example x ≥ 3.
    2. Swap x and y everywhere: in the equation and in the note. The note x ≥ 3 becomes y ≥ 3, because y now stands for the old inputs.
    3. Solve for y. Undo the outside steps first, in backward order: the number added or subtracted, then the number multiplying, until the squared part stands alone.
    4. Take the square root of both sides and write ± in front of the root. Both signs are candidates at this point.
    5. Choose the sign by reading the note. If the note says y ≥ h, keep the plus sign. If it says y ≤ h, keep the minus sign.
    6. Finish getting y alone, then rename y as f⁻¹(x).
    7. State the domain of f⁻¹. It is the range of f: the outputs of f are the inputs of f⁻¹.
    8. Check with one number that is not the vertex. Put it through f, put the result through f⁻¹, and confirm you are back where you started.
  • Inverses of other polynomials
    1. Look at the shape. If x appears once, inside a single cube, the function is x3 shifted, stretched or flipped. It is one-to-one and needs no cut.
    2. Write y in place of f(x).
    3. Swap x and y. The equation now describes the inverse.
    4. Undo whatever sits outside the cube, last thing done first: move an added or subtracted number to the other side, then divide or multiply away the number in front.
    5. With the cube alone on its side, take the cube root of both sides. Write no ± sign.
    6. If anything is still attached to y, such as y − 2, undo it. That number lands outside the root.
    7. Rename y as f−1(x). State the domain and the range: both are all real numbers.
    8. Check with a number. Put an input into f, feed the output into f−1, and confirm that the input comes back.
  • Radical functions and their domains
    1. Find the root sign and name it: square root (even) or cube root (odd). These steps are for an inside of the form ax + b with a ≠ 0.
    2. If it is a cube root, stop: a cube root accepts every real number, so the domain is every real number, (−∞, ∞).
    3. If it is a square root, copy everything under the root sign and write: inside ≥ 0.
    4. Solve the inequality for x the way you solve an equation: undo the adding or subtracting first, then undo the multiplying. If you multiply or divide both sides by a negative number, flip the sign.
    5. Write the domain as an inequality and as an interval. When the root is not the bottom of a fraction, the endpoint gets a square bracket, because an inside of 0 is allowed: 0 = 0.
    6. Check with three numbers: the endpoint (the inside should be 0), one number in the domain (the inside should be positive), and one number outside it (the inside should be negative).
  • Inverses of rational functions
    1. Find the input f cannot take: set the bottom of f equal to 0 and solve. Leave that number out of the domain of f.
    2. Write y in place of f(x).
    3. Swap x and y everywhere.
    4. Multiply both sides by the whole bottom, written in brackets, so the fraction disappears.
    5. Distribute to remove the brackets.
    6. Gather: move every term that contains y to one side and every term without y to the other side.
    7. Factor y out of the side that holds it, so y appears once, times a bracket.
    8. Divide both sides by that bracket. Rename y as f⁻¹(x).
    9. Find the input f⁻¹ cannot take: set the bottom of f⁻¹ equal to 0 and solve. That number is left out of the domain of f⁻¹ and out of the range of f.
    10. Trade: the range of f⁻¹ is the domain of f from step 1.
    11. Check with one number: pick an input, run f, feed the output to f⁻¹, and confirm you get the input back.
  • A function and its inverse undo each other
    1. Write the domain beside each function, including any cut. The domains decide the last step of the check.
    2. Build f(g(x)): copy the rule of f, and wherever f has x, write the whole rule of g in parentheses.
    3. Tidy up. Squaring undoes a square root, cubing undoes a cube root, and opposite steps such as + 3 and − 3 cancel. The result must be exactly x, with nothing left over.
    4. Build g(f(x)) the same way, with the whole rule of f inside g, and tidy up.
    5. If a square root of a square appears, such as x2, write it as |x| first. Then use the domain to decide whether |x| is x or −x.
    6. If both orders give x, the two functions are inverses. Say where each result holds: f(g(x)) = x on the domain of g, and g(f(x)) = x on the domain of f.
    7. Check with a number. Pick an allowed input a, compute b = f(a), and confirm that g(b) = a. The points (a, b) and (b, a) are mirror images across y = x.
Traps (6)
  • Why a quadratic needs its domain cut to have an inverse

    Cutting at x = 0 out of habit. That works only when the vertex sits on the y-axis, as in x2 or x2 + 5. For (x − 4)2, 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.

  • Finding the inverse of a quadratic on a cut domain

    Writing the plus root out of habit. The root step always produces ±, and the plus sign is right only when the right half of the parabola was kept. When the restricted domain is x ≤ h, the answer needs the minus sign. To avoid the mistake, carry the domain note through the swap, read it before you choose the sign, and test one number that is not the vertex.

  • Inverses of other polynomials

    The most common mistake is carrying square root habits into cube roots: writing ± in front of the cube root, or cutting the domain as if the cubic were a parabola. A cube root has one value. 83 is 2 only, because (−2)3 is −8 and not 8. Before writing ±, ask whether the power being undone is even or odd. Only an even power loses the sign.

  • Radical functions and their domains

    Writing x ≥ 0 out of habit. The condition belongs to the whole inside, not to x alone. For f(x) = 2x+6, the answer x ≥ 0 throws away every input from −3 up to, but not including, 0, and those inputs work: x = −1 gives 4 = 2. To avoid it, copy everything under the root sign first, write ≥ 0 after it, and only then solve.

  • Inverses of rational functions

    After the fraction is cleared, y sits in two terms on opposite sides, as in xy − 3x = 2y + 1. The common mistake is to divide right away and write y = 2y+3x+1x. That is not solved, because y is still on the right. Gather every y term on one side first, factor y out, and only then divide.

  • A function and its inverse undo each other

    Checking one order only, or dropping the domain. With f(x) = x2 on all real numbers and g(x) = x, the order f(g(x)) = (x)2 = x works, yet g(f(−3)) = 9 = 3, not −3. Always work both orders, write x2 as |x|, and write the domain beside each result.

Lesson 1 of 6

Why a quadratic needs its domain cut to have an inverse

Subject: MathematicsLeans on: One-to-one functionLeans on: Inverse functionLeans on: Parabola and vertexLeans on: Vertex from ax² + bx + cLeans on: Opens up or opens downLeans on: Squaring a negative numberLeans on: Order of operationsLeans on: Interval notationOpen this lesson
Leans on 1Comes back in 0

1 In plain words

Picture a machine that squares numbers. Put in 3 and 9 comes out. Put in −3 and 9 comes out again. Now run the machine backward: hand it 9 and ask what went in. It cannot answer, because two different numbers made 9. An inverse function is that backward machine, and it must give one clear answer every time. The fix is to agree ahead of time to use only half of the inputs, for example only 0 and up. Then 9 can only have come from 3. On the graph, you keep one half of the U-shaped curve, split at its turning point, the vertex. The inputs you keep are called the restricted domain.

2 Picture it

(0, 0)(−2, 4)(2, 4)vertex (0, 0)−4−224246810(0, 0)(−2, 4)(2, 4)vertex (0, 0)−4−3−2−11234246810
The horizontal line at height 4 meets y = x2 twice, once on each side of the vertex.

Reminder

Reminder: One-to-one function, Inverse function, Parabola and vertex, Vertex from ax2 + bx + c, Opens up or opens down, Squaring a negative number, Order of operations, Interval notation
  • One-to-one function. Every output comes from exactly one input. On a graph, no horizontal line meets it twice. y = 2x + 1 passes. y = x2 fails: the line at height 4 meets it at x = −2 and at x = 2.
  • Inverse function. The function that undoes another one. If f sends 3 to 9, its inverse function sends 9 back to 3.
  • Parabola and vertex. The graph of a quadratic is a U-shaped curve called a parabola, and its turning point is the vertex. For a(x − h)2 + k the vertex is (h, k), with the opposite sign inside: (x + 3)2 − 2 is (x − (−3))2 − 2, so the vertex is (−3, −2).
  • Vertex from ax2 + bx + c. h = −b2a, then k = f(h). For x2 − 6x + 1: h = −−62 = 3 and k = 9 − 18 + 1 = −8.
  • Opens up or opens down. In a(x − h)2 + k, a positive a opens the parabola upward and the vertex is the lowest point. A negative a opens it downward and the vertex is the highest point.
  • Squaring a negative number. A negative times a negative is positive: (−3)2 = (−3) × (−3) = 9, the same as 32.
  • Order of operations. Square first, then multiply, then add: −2(0 − 1)2 + 8 = −2 × 1 + 8 = 6.
  • Interval notation. x ≥ 2 is written [2, ∞). The square bracket says 2 is included, and infinity always gets a round one. x ≤ −3 is written (−∞, −3].

3 Why

A parabola is a mirror image of itself across the vertical line through its vertex. Step the same distance d to the left and to the right of the vertex, to h − d and h + d, and the squared part is (−d)2 on one side and d2 on the other. Those are equal, so the two outputs are equal. Every output except the vertex's own comes from two inputs, one on each side. Keeping one side keeps exactly one input from each pair. Cut anywhere else and you keep both inputs of some pair, or lose outputs for no reason.

4 The rule

A quadratic f(x)=a(x−h)2+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.

The same idea, five ways

Say it

f of x equals x minus 2, squared, plus 1, for x at least 2. Or say: f, restricted to x greater than or equal to 2.

Write it

The function is used only for inputs on one side of the vertex, here the inputs that are 2 or more.

In math
  • f(x)=(x−2)2+1,x≥2
  • domain: [2, ∞)
  • {x | x ≥ 2}
  • the left half instead: x ≤ 2, or (−∞, 2]
  • graph words: the right half of the parabola, starting at the vertex (2, 1)
Like

A road that dips into a valley and climbs out passes each height twice. Keep only the climb out, and each height marks one spot.

See it
2[2, ∞)
The restricted domain x ≥ 2: a filled dot at 2, because the vertex stays, and shading to the right.

The same idea, other ways

As a picture

Look at the right half of y = x2, drawn here starting at the vertex (0, 0). The horizontal line at height 4 meets it once, at (2, 4). On the full parabola the same line also meets the left half, at (−2, 4). Erasing the left half is what cutting the domain means. The curve that is left passes the horizontal line test, so it is the graph of a one-to-one function.

(0, 0)(2, 4)vertex (0, 0)domainrange24246810
Only the right half is drawn: the line at height 4 now meets the curve once.
As a table

In the table for y = x2, the output 4 sits in two cells: under −2 and under 2. The arrows point up to both inputs. That is one output with two inputs, so the full table does not describe a one-to-one function. Now cover every column to the left of 0. In what is left, each output appears once.

input xoutput x²−39−24−1100112439↑ solve: output given, read every input above it
The two highlighted 4s in the output row come from the inputs −2 and 2, one on each side of 0.
As a story

You hike down into a valley and up the other side. Your altimeter reads 500 feet twice: once on the way down and once on the way up. A friend who hears only the number 500 cannot tell where you are on the trail. If you both agree first that you are on the uphill half, the height alone pins down your spot. The lowest point of the valley is the vertex, and the uphill half is the restricted domain.

Run it backward

An inverse function sends each output back to its input. For y = x2 on all real numbers, the output 9 would have to go back to 3 and also to −3. One starting value with two arrows is not a function at all. Keep only x ≥ 0 and the arrow to −3 disappears. Now 9 goes back to 3 and to nothing else.

93-3not a function
Running y = x2 backward, the single value 9 points to two answers.
As a rule of thumb

Fold at the vertex, keep one side. Find the x value of the vertex, h. Keep all of x ≥ h or all of x ≤ h. A domain with inputs on both sides of h works only if no two of them sit the same distance from h, so the usual cut keeps one whole side. If the problem does not say which side, keep the right side.

FunctionVertexRight halfLeft halfRange
x2(0, 0)x ≥ 0x ≤ 0y ≥ 0
(x−5)2(5, 0)x ≥ 5x ≤ 5y ≥ 0
(x+3)2−2(−3,−2)x≥−3x≤−3y≥−2
−2(x−1)2+8(1, 8)x ≥ 1x ≤ 1y ≤ 8

5 How

  1. Find the vertex (h, k). In the form a(x − h)2 + k, read it off, remembering that the sign inside the parentheses is the opposite of h. In the form ax2 + bx + c, compute h = −b2a, then k = f(h).
  2. 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.
  3. Choose one half. If the problem names a side, use that side. If it does not, keep the right half.
  4. 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.
  5. 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.
  6. 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.

Cutting a parabola so each output comes from one input

1
Is the quadratic written as a(x − h)2 + k?
YesRead the vertex (h, k) off it. The sign inside the parentheses is the opposite of h.
NoCompute h = −b2a, then k = f(h).
↓
2
Does the problem already give a domain or name a side?
YesCheck that it stays on one side of x = h. If it does, use it. If it reaches past h, move the cut to h.
NoKeep the right half: x ≥ h.
↓
3
Does the parabola open up (a > 0)?
YesThe vertex is the lowest point. For a kept side that runs on without end, the range is y ≥ k when it begins at x = h, and y ≥ f(c) when it begins at another x, c. For a stretch with two ends, the range runs between the outputs at the two ends, each one included only if its input is kept.
NoThe vertex is the highest point. For a kept side that runs on without end, the range is y ≤ k when it begins at x = h, and y ≤ f(c) when it begins at another x, c. For a stretch with two ends, the range runs between the outputs at the two ends, each one included only if its input is kept.
↓
4
Are two allowed inputs the same distance from h on opposite sides?
YesThey give the same output, so the cut is in the wrong place. Move it to the vertex.
NoEach output comes from one input. The cut function is a one-to-one function and has an inverse function.
  1. Find the vertex (h, k).
  2. Pick the side to keep: the one the problem names, or the right side if it names none.
  3. Write the restricted domain: x ≥ h or x ≤ h, with the vertex included.
  4. Write the range from k: upward if the parabola opens up, downward if it opens down.
  5. Test two inputs that mirror each other across x = h. Only one of them may still be allowed.

6 Worked example

Worked example

Cutting (x − 2)2 + 1 at its vertex

Let f(x) = (x − 2)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.

Show the steps and the answer
(2, 1)(0, 5)(4, 5)vertex (2, 1)24246810
Before the cut: the line at height 5 meets the parabola at x = 0 and at x = 4.
(2, 1)(4, 5)vertex (2, 1)24246810
After the cut: only the half with x ≥ 2 is drawn, and the line at height 5 meets it once.
input xoutput f(x)213245510↑ solve: output given, read every input above it
On the kept half the output 5 sits in one cell, under the input 4.
  1. Read the vertex off the formula. f(x)=(x−2)2+1 matches a(x−h)2+k with a=1,h=2 and k=1. The vertex is (2, 1).The squared part (x − 2)2 is smallest, namely 0, when x = 2, and then f(2) = 0 + 1 = 1. The sign inside the parentheses is the opposite of h: x − 2 means h = 2.
  2. Pick two inputs the same distance from 2, one on each side: x=0 and x=4. Each is 2 units from the vertex.The parabola is a mirror image across the line x = 2, so a mirrored pair like this is where a repeated output shows up.
  3. Compute both outputs. f(0)=(0−2)2+1=(−2)2+1=4+1=5. f(4)=(4−2)2+1=22+1=4+1=5.Squaring erases the sign: (−2)2 and 22 are both 4.
  4. Conclude that f on all real numbers is not a one-to-one function: the inputs 0 and 4 are different, and both give 5.A one-to-one function never sends two inputs to the same output. On the first graph, the horizontal line at height 5 meets the parabola at (0, 5) and at (4, 5).
  5. Cut at the vertex and keep the right half. The restricted domain is x≥2, which is [2,∞) in interval notation.To the right of the vertex the curve only climbs, so no height can come up twice. The 2 itself stays in, because the output 1 is made by the input 2 alone.
  6. Look again at the pair from step 3. The input 0 is less than 2, so it is no longer allowed. The input 4 is still allowed. The output 5 now comes from 4 and from nothing else.The cut removes exactly one input from every mirrored pair. The second graph shows the line at height 5 meeting the kept half once.
  7. Write the range: y≥1, which is [1,∞).a = 1 is positive, so the parabola opens up and the vertex is its lowest point. (x − 2)2 is never negative, so f(x) is never below 0 + 1 = 1, and the right half alone still climbs through every height from 1 upward.
Answer
  • 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.
Double-check A second way, with algebra instead of the picture. Ask which inputs give the output 5: solve (x − 2)2 + 1 = 5. Subtract 1 from both sides: (x − 2)2 = 4. Two numbers square to 4, so x − 2 = 2 or x − 2 = −2, which gives x = 4 or x = 0. On all real numbers that is two inputs. On x ≥ 2 only x = 4 is allowed, so it is one input. The table of the kept half agrees: read along its output row and the numbers only rise, with none appearing twice.

Ladder: from easy to exam-hard

Try each rung before you open its steps.

Rung 1 of 6

The plain parabola

f(x) = x2. Cut the domain so that f is a one-to-one function, keeping the right half. Give the restricted domain and the range. In plain words: decide which inputs to keep so that no output is made twice.

Show the steps and the answer
input xoutput x²00112439↑ solve: output given, read every input above it
With inputs from 0 up, the output 9 appears once, under the input 3.
  1. Find the vertex. x2 is the same as 1(x−0)2+0, so h=0 and k=0. The vertex is (0, 0).x2 is smallest, namely 0, when x = 0.
  2. See the problem first: f(−3)=(−3)2=9 and f(3)=32=9.Two different inputs with the same output means f on all real numbers is not a one-to-one function.
  3. Keep the right half: x≥0, or [0,∞).From every pair such as −3 and 3, only the one at or to the right of the vertex stays.
  4. Write the range: y≥0, or [0,∞).A square is never negative, and the right half still reaches every height from 0 up.
Answer
  • Restricted domain: x ≥ 0, or [0, ∞).
  • Range: y ≥ 0, or [0, ∞).
Double-check Look at the table of the kept half. Read along the output row: the numbers only rise, and none appears twice. The output 9 sits under the input 3 alone, because −3 is no longer an allowed input.

Rung 2 of 6

The vertex moved sideways

f(x) = (x − 5)2. Cut the domain so that f is a one-to-one function, keeping the right half. Give the restricted domain and the range. In plain words: find where the parabola turns, then keep the inputs from there on.

Show the steps and the answer
input xoutput f(x)50617489
Starting at the vertex input 5, every output appears once.
  1. Find the vertex. (x−5)2 matches (x−h)2 with h=5, and nothing is added, so k=0. The vertex is (5, 0).The square is 0 when x = 5. The sign inside is the opposite of h: x − 5 means h = 5.
  2. Test a mirrored pair: f(4)=(4−5)2=(−1)2=1 and f(6)=(6−5)2=12=1.4 and 6 are each 1 unit from the vertex, on opposite sides, so they give the same output.
  3. Keep the right half: x≥5, or [5,∞).The cut goes at the x value of the vertex, which is 5, not at 0. Now 4 is out and 6 stays.
  4. Write the range: y≥0, or [0,∞).The parabola opens up and its lowest output is f(5) = 0.
Answer
  • Restricted domain: x ≥ 5, or [5, ∞).
  • Range: y ≥ 0, or [0, ∞).
Double-check In the table of the kept half, the output row rises from 0 with no number repeated. Test the tempting cut as well: with x ≥ 0 the inputs 4 and 6 would both be allowed, and both give 1, so that cut fails.

Rung 3 of 6

A plus sign inside, and the left half

f(x) = (x + 3)2 − 2. Cut the domain so that f is a one-to-one function, this time keeping the left half. Give the restricted domain and the range. In plain words: find the turning point, then keep the inputs at or to the left of it.

Show the steps and the answer
−3(−∞, −3]
The restricted domain x ≤ −3: a filled dot at −3 and shading to the left.
input xoutput f(x)−67−52−4−1−3−2↑ solve: output given, read every input above it
On the left half the output −1 appears once, under the input −4.
  1. Rewrite x+3 as x−(−3). So h=−3 and k=−2. The vertex is (−3,−2).The form a(x − h)2 + k has a subtraction inside, so a plus 3 means h = −3. Plug it in to confirm: f(−3) = (−3 + 3)2 − 2 = 0 − 2 = −2.
  2. Test a mirrored pair: f(−4)=(−4+3)2−2=(−1)2−2=1−2=−1, and f(−2)=(−2+3)2−2=12−2=1−2=−1.−4 and −2 are each 1 unit from −3, on opposite sides.
  3. Keep the left half: x≤−3, or (−∞,−3].The left half is the inputs at or to the left of the vertex. The square bracket sits on −3 because the vertex stays in.
  4. Write the range: y≥−2, or [−2,∞).a = 1 is positive, so the parabola opens up and its lowest output is k = −2. Moving left from the vertex, the left half climbs through every height above −2.
Answer
  • Restricted domain: x ≤ −3, or (−∞, −3].
  • Range: y ≥ −2, or [−2, ∞).
Double-check In the table of the kept half, no number appears twice in the output row. The output −1 from step 2 sits under the input −4 alone, because −2 is to the right of the vertex and is not an allowed input.

Rung 4 of 6

A parabola that opens down

f(x) = −2(x − 1)2 + 8. Cut the domain so that f is a one-to-one function, keeping the right half. Give the restricted domain and the range. In plain words: the curve is an upside-down U, so find its top, keep the inputs from there on, and say which heights it reaches.

Show the steps and the answer
(1, 8)vertex (1, 8)(2, 6)24−8−448
The kept right half starts at the top, (1, 8), and only falls.
input xoutput f(x)1826304−10↑ solve: output given, read every input above it
The output 6 appears once, under the input 2.
  1. Read the vertex: a=−2,h=1,k=8. The vertex is (1, 8).The formula matches a(x − h)2 + k. Plug in to confirm: f(1) = −2 × 02 + 8 = 8.
  2. Notice that a=−2 is negative, so the parabola opens down and the vertex is the highest point.−2 times a square is never positive, so f(x) is never above 8.
  3. Test a mirrored pair: f(0)=−2(0−1)2+8=−2×1+8=6, and f(2)=−2(2−1)2+8=−2×1+8=6.Square first, then multiply by −2, then add 8. The inputs 0 and 2 are each 1 unit from 1, on opposite sides.
  4. Keep the right half: x≥1, or [1,∞).The cut still goes at the x value of the vertex. Opening down changes the range, not the place of the cut.
  5. Write the range: y≤8, or (−∞,8].The vertex is the top, and the right half falls through every height below 8.
Answer
  • Restricted domain: x ≥ 1, or [1, ∞).
  • Range: y ≤ 8, or (−∞, 8].
Double-check In the table of the kept half, the output row only falls and no number appears twice. The output 6 sits under the input 2 alone, because 0 is left of the vertex and is not allowed.

Rung 5 of 6

Standard form, so find the vertex first

f(x) = x2 − 6x + 1. Cut the domain so that f is a one-to-one function. Give the restricted domain and the range. In plain words: the turning point is hidden in this form, so compute it, then keep the inputs from there on.

Show the steps and the answer
(3, −8)(2, −7)(4, −7)vertex (3, −8)246−8−6−4−22
The line at height −7 meets the full parabola at x = 2 and x = 4, one on each side of the vertex.
input xoutput f(x)3−84−75−461
From the vertex input 3 onward, the outputs rise and none repeats.
  1. Name the numbers in ax2 + bx + c: a=1,b=−6,c=1.The formula for the vertex uses a and b.
  2. Compute h=−b2a=−−62×1=−−62=−(−3)=3.This finds the one input where the parabola turns around, which is where the cut goes.
  3. Plug 3 back in: k=f(3)=32−6×3+1=9−18+1=−8. The vertex is (3,−8).The output at the turning input is the height of the vertex.
  4. Test a mirrored pair: f(2)=22−6×2+1=4−12+1=−7, and f(4)=42−6×4+1=16−24+1=−7.2 and 4 are each 1 unit from 3. Equal outputs confirm that x = 3 is the mirror line, and that f on all real numbers is not a one-to-one function.
  5. Keep the right half: x≥3, or [3,∞).The problem names no side, so take the usual choice, the right half.
  6. Write the range: y≥−8, or [−8,∞).a = 1 is positive, so the parabola opens up and its lowest output is k = −8.
Answer
  • Vertex: (3, −8).
  • Restricted domain: x ≥ 3, or [3, ∞).
  • Range: y ≥ −8, or [−8, ∞).
Double-check In the table of the kept half, the output row rises from −8 and no number appears twice. The vertex is confirmed a second way by step 4: the outputs at 2 and at 4 match, and f(3) = −8 is lower than both.

Rung 6 of 6

Is the given cut in the right place?

g(x) = 2x2 + 8x + 3 is given with the domain x ≤ −1. Is g a one-to-one function on that domain? If not, move the cut so that it is, keeping the left side, and give the range. In plain words: check whether the allowed inputs stay on one side of the turning point. If they do not, find two allowed inputs with the same output and repair the domain.

Show the steps and the answer
(−2, −5)(−3, −3)(−1, −3)vertex (−2, −5)−4−2−44812
The inputs −3 and −1 both satisfy x ≤ −1 and both give −3, so the given cut is too far right.
input xoutput g(x)−513−43−3−3−2−5
With the cut moved to x ≤ −2, every output appears once.
  1. Name a=2 and b=8, then compute h=−b2a=−82×2=−84=−2.This finds the input where the parabola turns around, the only place a cut works without losing outputs.
  2. Plug −2 back in: k=g(−2)=2×(−2)2+8×(−2)+3=8−16+3=−5. The vertex is (−2,−5).The output at the turning input is the height of the vertex.
  3. Compare the given domain with the vertex. x≤−1 reaches past −2, so it holds inputs on both sides of the vertex, for example −3 and −1.An interval domain works only if it stays on one side of x = h. Here −3 is left of −2 and −1 is right of −2, and both satisfy x ≤ −1.
  4. Compute both outputs: g(−3)=2×9+8×(−3)+3=18−24+3=−3, and g(−1)=2×1+8×(−1)+3=2−8+3=−3.−3 and −1 are each 1 unit from −2, so they mirror each other. Two allowed inputs with the same output means g is not a one-to-one function on x ≤ −1.
  5. Move the cut to the vertex: x≤−2, or (−∞,−2].Now −1 is out and −3 stays, so the output −3 comes from one input.
  6. Write the range: y≥−5, or [−5,∞).a = 2 is positive, so the parabola opens up and its lowest output is k = −5.
Answer
  • No. On x ≤ −1, g(−3) = −3 and g(−1) = −3.
  • Corrected restricted domain: x ≤ −2, or (−∞, −2].
  • Range: y ≥ −5, or [−5, ∞).
Double-check In the table of the corrected domain, no number appears twice in the output row, and the output −3 sits under the input −3 alone. The graph shows the same thing: the line at height −3 meets the full parabola on both sides of x = −2, and only the left meeting point has x ≤ −2.

Counterexamples

What it is not, and the tempting wrong moves.

✗ Not this: For f(x) = (x − 4)2, use the restricted domain x ≥ 0.
The vertex is at x = 4, not 0. The inputs 3 and 5 are both allowed and f(3) = (−1)2 = 1 and f(5) = 12 = 1. Two inputs still share an output.
✓ Instead: Cut at the vertex: x ≥ 4 for the right half, or x ≤ 4 for the left half.
✗ Not this: Any smaller domain will do. For f(x) = x2, use −1 ≤ x ≤ 3.
That domain still has inputs on both sides of the vertex at 0. f(−1) = 1 and f(1) = 1. Making the domain smaller is not enough. An interval domain has to stay on one side of the vertex.
✓ Instead: Use x ≥ 0 or x ≤ 0. The shorter stretch 0 ≤ x ≤ 3 would also be one-to-one, but it loses every output above 9.
✗ Not this: For f(x) = (x + 3)2 − 2, read the vertex as x = 3 and keep x ≥ 3.
x + 3 means h = −3. The domain x ≥ 3 does give a one-to-one function, but it starts at f(3) = 36 − 2 = 34, so every output from −2 up to, but not including, 34 is thrown away for no reason.
✓ Instead: Keep x ≥ −3. The range is then all of y ≥ −2, the same as the whole parabola.
✗ Not this: For f(x) = (x − 2)2 + 1 with x ≥ 2, write the range as y ≥ 2.
That mixes up h and k. The 2 is the input at the vertex. The output there is f(2) = 0 + 1 = 1.
✓ Instead: Domain uses h: x ≥ 2. Range uses k: y ≥ 1.
✗ Not this: Cut the domain of f(x) = 2x + 1 before looking for its inverse function.
This is a case where the rule does not apply. A slanted line already passes the horizontal line test, so no two inputs share an output. Cutting it only discards inputs.
✓ Instead: Cut a domain only when the function fails the horizontal line test, as every parabola does.

Tips and tricks

  • Memory cue: fold at the vertex, keep one side.
  • Quick check for any cut: h − 1 and h + 1 always give the same output, so a domain that allows both is not one-to-one. Passing this check is not proof: an interval domain must also lie wholly on one side of h.
  • The vertex stays in the half you keep. Write x ≥ h or x ≤ h with a square bracket, because the output k comes from the input h alone.
  • Cutting at the vertex never changes the range. Each half reaches every height the whole parabola reaches.
  • Domain uses h, range uses k. Say it once before writing the answer: inputs from h, outputs from k.
  • When no side is named, keep the right half. When a side is named, or the function comes with a domain, use it and check that it stays on one side of h.
  • You do not need to memorize the table of cuts. Find the vertex and rebuild the answer in a few seconds.

7 The trap

Cutting at x = 0 out of habit. That works only when the vertex sits on the y-axis, as in x2 or x2 + 5. For (x − 4)2, 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.

Flash cards

Say the answer out loud, then flip.

A machine squares numbers. Why can it not be run backward when every number is allowed in?
  • Two different inputs give the same output: 3 and −3 both give 9.
  • Handed 9, the backward machine has two possible answers. An inverse function must give one clear answer every time.
Why is x ≥ 0 the wrong cut for f(x) = (x − 6)2?
  • The vertex is at x = 6, not at 0.
  • The inputs 5 and 7 are both allowed by x ≥ 0, and both give 1, so nothing is fixed.
  • Cut at the vertex: x ≥ 6 or x ≤ 6.
f(x) = x2 − 6x + 2. What restricted domain keeps the right half of the parabola?
  • h = −b2a = −−62 = 3, so the restricted domain is x ≥ 3, which is [3, ∞).
  • Check: f(2) = −6 and f(4) = −6. Same output, and only 4 is still allowed.
You cut a parabola at its vertex (h, k) and keep one half. Does the range change?
  • No. One half still reaches every height the whole curve reached.
  • If a > 0 it opens up: y ≥ k.
  • If a < 0 it opens down: y ≤ k.

Lesson 2 of 6

Finding the inverse of a quadratic on a cut domain

Subject: MathematicsLeans on: The square root step and its signLeans on: The symbol √ is never negativeLeans on: Same move on both sidesLeans on: Undo in backward orderLeans on: A square is never negativeLeans on: The domain of a square rootLeans on: Flipping an inequalityLeans on: One-to-one functionOpen this lesson
Leans on 0Comes back in 0

1 In plain words

Think about putting on socks, then shoes. To undo that, you take off the shoes first, then the socks: the same steps, in backward order, each one reversed. Finding an inverse function works the same way. If f squares an input and then adds 3, the inverse subtracts 3 and then undoes the square. Undoing the square is the one step that needs care. Suppose a friend says: I am 2 blocks from your home. That does not say east or west. Undoing a square has the same problem: a square above 0 comes from two numbers, one plus and one minus. The restricted domain from the last lesson is the missing direction. It tells you which half of the parabola was kept, and that picks the sign.

2 Picture it

(2, 3)(6, 5)(6, 1)24681012246(2, 3)(6, 5)(6, 1)1357911246
The solid curve is 3 + x−2 and the dashed curve is 3 − x−2.

Reminder

Reminder: The square root step and its sign, The symbol √ is never negative, Same move on both sides, Undo in backward order, A square is never negative, The domain of a square root, Flipping an inequality, One-to-one function
  • The square root step and its sign. If a square equals 9, the thing squared is 3 or −3. Write this as ±9, read plus or minus root 9.
  • The symbol √ is never negative. 9 means 3 only. To get −3 you must write −9.
  • Same move on both sides. An equation stays true when both sides get the same change: x = y2 + 5 becomes x − 5 = y2 after subtracting 5 from both sides.
  • Undo in backward order. To undo times 2 then add 3, subtract 3 first, then divide by 2. From 2y + 3 = 11: 2y = 8, then y = 4.
  • A square is never negative. (x − 3)2 is 0 or more for every x, so (x − 3)2 + 2 is 2 or more. That gives the range of f.
  • The domain of a square root. The inside of a square root must be 0 or more: x−2 needs x − 2 ≥ 0, so x ≥ 2.
  • Flipping an inequality. Multiplying or dividing both sides by a negative number flips the sign: −x ≥ −8 becomes x ≤ 8.
  • One-to-one function. Each output comes from one input only. x2 on x ≥ 0 is one: the output 9 comes from 3 and from nothing else.

3 Why

A function sends an input x to an output y. The inverse function runs the other way: its input is the old output and its output is the old input. Swapping the letters x and y in the equation writes exactly that trade, and solving for y puts the new output alone. The square root step gives two candidates because two numbers share each square above 0: 32 = 9 and (−3)2 = 9. Only the square 0 comes from one number. Only one candidate lies in the restricted domain, so only one sign returns the input you started with. The other sign is the inverse of the half that was cut away.

4 The rule

For f(x)=a(x−h)2+k on the restricted domain x≥h, the inverse function is f−1(x)=h+(x−k)÷a.

On the restricted domain x≤h it is f−1(x)=h−(x−k)÷a. The domain of f−1 is the range of f, and the range of f−1 is the restricted domain of f.

The same idea, five ways

Say it

Say f⁻¹(x) as: f inverse of x. Say the method as: swap, solve, choose the sign.

Write it

The inverse function of a quadratic on its restricted domain is the rule that takes each output of the quadratic back to the one input, on the kept half, that made it.

In math
  • f⁻¹(x)
  • f(a) = b means f⁻¹(b) = a
  • y = (x − h)2 + k with x ≥ h becomes x = (y − h)2 + k with y ≥ h
  • f⁻¹(x) = h + x−k on [k, ∞) when the right half was kept
  • f⁻¹(x) = h − x−k on [k, ∞) when the left half was kept
Like

Socks, then shoes on the way out. Shoes, then socks on the way back. The same steps, backward, each one reversed.

See it

Start: y = (x − 3)2 + 2, x ≥ 3

Swap: x = (y − 3)2 + 2, y ≥ 3

Root: y − 3 = ±x−2

Note says y ≥ 3, so keep +

f⁻¹(x)=3+x−2,x≥2

The same idea, other ways

As a story

Your home is at 3 on a number line. Someone says: I am 2 blocks from your home. That could be 5 or it could be 1. The square in (y − 3)2 does the same thing: its value depends only on how far y is from 3, not on which way. If you know the person only walks east of home, which is y ≥ 3, the answer is 3 + 2 = 5. If the person only walks west, which is y ≤ 3, the answer is 3 − 2 = 1. The restricted domain is the direction.

135home
Both 1 and 5 sit 2 steps from home at 3, but only 5 is on the shaded side that was kept.
With small numbers

Take f(x) = x2. Both 3 and −3 square to 9, so the output 9 has two inputs and cannot be sent back to one place. Keep only x ≥ 0, and the input was 3: the inverse is x, because 9 = 3. Keep only x ≤ 0, and the input was −3: the inverse is −x, because −9 = −3. The formula for f is the same both times, yet the inverses differ. The restricted domain is the only thing that tells them apart.

−339function
Two inputs, −3 and 3, land on the same output 9, so the trip back must know which half was kept.
As a table read backward

The table shows f(x) = (x − 3)2 + 2 on x ≥ 3. To read f, start in the top row and go down. To read the inverse, start in the bottom row and go up: find 6 in the output row, and the input above it is 5. So f⁻¹(6) = 5. The formula 3 + x−2 does the same job without a table: 3 + 6−2 = 3 + 2 = 5.

input xoutput f(x)324356611↑ solve: output given, read every input above it
The up arrow from the output 6 to the input 5 is the inverse at work: f⁻¹(6) = 5.
Why it must be true

After the swap you reach (y − 3)2 = x − 2, so y − 3 is either x−2 or −x−2. The symbol √ always gives 0 or more. If the note says y ≥ 3, then y − 3 is 0 or more, so +x−2 fits for every x ≥ 2, where the root is defined; −x−2 fits only at x = 2, where both signs give 0. If the note says y ≤ 3, then y − 3 is 0 or less, so −x−2 fits for every x ≥ 2. One note allows one formula. Nothing is left to guess.

As a rule of thumb

Right half, plus root. Left half, minus root. The right half of a parabola holds the inputs at or to the right of the vertex's x value h, so the inverse must add the root to reach them. The left half holds the inputs at or to the left of h, so the inverse must subtract the root. If you forget which is which, test one number: the wrong sign sends you to the half that was cut away.

Right half kept (x ≥ h): h + ...

Left half kept (x ≤ h): h − ...

Unsure? Test a number that is not the vertex.

Half keptNote after the swapSign of the rootExample with (x − 3)2 + 2
Right half: x ≥ hy ≥ h, so y − h is 0 or more+On x ≥ 3: f⁻¹(x) = 3 + x−2
Left half: x ≤ hy ≤ h, so y − h is 0 or less−On x ≤ 3: f⁻¹(x) = 3 − x−2

5 How

  1. Write y in place of f(x), and write the restricted domain (the cut domain) beside it as a note, for example x ≥ 3.
  2. Swap x and y everywhere: in the equation and in the note. The note x ≥ 3 becomes y ≥ 3, because y now stands for the old inputs.
  3. Solve for y. Undo the outside steps first, in backward order: the number added or subtracted, then the number multiplying, until the squared part stands alone.
  4. Take the square root of both sides and write ± in front of the root. Both signs are candidates at this point.
  5. Choose the sign by reading the note. If the note says y ≥ h, keep the plus sign. If it says y ≤ h, keep the minus sign.
  6. Finish getting y alone, then rename y as f⁻¹(x).
  7. State the domain of f⁻¹. It is the range of f: the outputs of f are the inputs of f⁻¹.
  8. Check with one number that is not the vertex. Put it through f, put the result through f⁻¹, and confirm you are back where you started.

Finding the inverse of a quadratic on a restricted domain

1
Does the quadratic come with a restricted domain that makes it a one-to-one function?
YesGo on.
NoStop. Cut the domain at the vertex first, as in the last lesson.
↓
2
After the swap, is a number added or subtracted outside the square?
YesUndo it first, on both sides.
NoGo on.
↓
3
Is a number multiplying the squared part?
YesDivide both sides by it before taking the root.
NoTake the root now.
↓
4
Does the note say y ≥ h, meaning the right half was kept?
YesKeep the plus sign in front of the root.
NoThe note says y ≤ h, the left half. Keep the minus sign.
↓
5
Does your check number come back to itself?
YesDone. Write the inverse with its domain.
NoLook at the sign of the root first, then at the order of the undo steps.
  1. Write y = the formula, and beside it the restricted domain as a note.
  2. Swap x and y in the equation and in the note.
  3. Undo the outside steps in backward order until the squared part stands alone.
  4. Take the square root of both sides and write ±.
  5. Read the note to choose the sign: y ≥ h keeps +, y ≤ h keeps −.
  6. Finish solving for y and rename it f⁻¹(x).
  7. State the domain of f⁻¹: it is the range of f.
  8. Check with one number that is not the vertex: through f, then through f⁻¹, and back.

6 Worked example

Worked example

A shifted parabola with the right half kept

Find the inverse function of f(x) = (x − 3)2 + 2 on the restricted domain x ≥ 3, and state the domain of the inverse.

Show the steps and the answer
5f: subtract 3,square, add 2f⁻¹: subtract2, take theroot, add 3firstsecond65
The number 5 goes through f and comes out as 6, then goes through f⁻¹ and comes back as 5.
  1. Say what is asked. f takes an input that is 3 or more and gives an output. You want the rule that takes each output back to the input that made it.That is what an inverse function does: it runs f backward.
  2. Write y=(x−3)2+2, with the note x≥3.y is a shorter name for the output f(x), and the note keeps the restricted domain in view.
  3. Swap x and y: x=(y−3)2+2, with the note y≥3.The inverse trades the roles of input and output. The note swaps too, since y now stands for the old inputs.
  4. Subtract 2 from both sides: x−2=(y−3)2.Adding 2 was the last thing f did, so it is the first thing to undo. This leaves the squared bracket alone.
  5. Take the square root of both sides: y−3=±x−2.Two numbers have the same square, so the root step must offer both signs for now.
  6. Choose the sign. The note says y ≥ 3, so y − 3 is 0 or more. Keep the plus sign: y−3=x−2.The symbol √ never gives a negative number, so the plus root matches a quantity that is 0 or more. The minus root would make y − 3 negative for every x above 2, which the note forbids.
  7. Add 3 to both sides: y=3+x−2. Rename y: f−1(x)=3+x−2.y is now alone, and y is the output of the inverse.
  8. State the domain of the inverse: x≥2.The inverse takes the outputs of f as its inputs. A square is 0 or more, so (x − 3)2 + 2 is 2 or more. The root agrees: x − 2 must not be negative.
Answer
  • f⁻¹(x)=3+x−2
  • Domain of f⁻¹: x ≥ 2, which is [2, ∞)
  • Range of f⁻¹: y ≥ 3, the restricted domain of f
Double-check Send a number on a round trip. Start with 5, which is in the restricted domain. f(5) = (5 − 3)2 + 2 = 4 + 2 = 6. Then f⁻¹(6) = 3 + 6−2 = 3 + 4 = 3 + 2 = 5. You are back at 5. The other sign fails: 3 − 4 = 1, and 1 is not in x ≥ 3. A second trip: f(7) = 42 + 2 = 18, and f⁻¹(18) = 3 + 16 = 3 + 4 = 7.

Ladder: from easy to exam-hard

Try each rung before you open its steps.

Rung 1 of 6

The plain square, right half kept

Find the inverse function of f(x) = x2 on the restricted domain x ≥ 0.

Show the steps and the answer
9f⁻¹: take the squareroot, plus signbecause x ≥ 0 waskeptinputoutput3
The inverse machine takes the output 9 back to 3, the input on the kept half.
  1. Say what is asked: f squares a number that is 0 or more. Find the rule that takes the square back to that number.The inverse function runs f backward.
  2. Write y=x2, with the note x≥0.y names the output, and the note records which half of the parabola was kept.
  3. Swap x and y: x=y2, with the note y≥0.The inverse trades input and output, so the letters trade places, in the note too.
  4. Take the square root of both sides: y=±x.A number and its negative have the same square, so both signs are candidates.
  5. Choose the sign. The note says y ≥ 0, so keep the plus sign: y=x.x is 0 or more, which matches the note. −x would be 0 or less.
  6. Write f−1(x)=x, with domain x≥0.The inverse takes the outputs of f as its inputs, and a square is never negative.
Answer
  • f⁻¹(x)=x
  • Domain of f⁻¹: x ≥ 0
Double-check Round trip with 4: f(4) = 16, then f⁻¹(16) = 16 = 4. Round trip with 7: f(7) = 49, then 49 = 7. Both come back.

Rung 2 of 6

One outside step to undo first

Find the inverse function of f(x) = x2 + 5 on the restricted domain x ≥ 0, and state its domain.

Show the steps and the answer
5[5, ∞)
The inverse accepts only 5 and above, because 5 is the smallest output f can make.
  1. Say what is asked: f squares an input that is 0 or more, then adds 5. Find the rule that goes from the result back to the input.The inverse must undo both steps, the last one first.
  2. Write y=x2+5, with the note x≥0. Swap: x=y2+5, with the note y≥0.Swapping the letters trades input and output, and the note follows the old inputs to their new letter.
  3. Subtract 5 from both sides: x−5=y2.Adding 5 was the last step of f, so it is undone first. The square now stands alone.
  4. Take the square root of both sides: y=±x−5.The root step always offers both signs before the note is read.
  5. Choose the sign. The note says y≥0, so y=x−5. Rename: f−1(x)=x−5.The plus root is 0 or more, which is what the note demands.
  6. State the domain: x≥5.x2 is 0 or more, so the outputs of f are 5 or more, and those are the inputs of the inverse. The root agrees: x − 5 must be 0 or more.
Answer
  • f⁻¹(x)=x−5
  • Domain of f⁻¹: x ≥ 5, which is [5, ∞)
Double-check f(2) = 4 + 5 = 9, then f⁻¹(9) = 9−5 = 4 = 2. Again with 6: f(6) = 36 + 5 = 41, then f⁻¹(41) = 41−5 = 36 = 6.

Rung 3 of 6

The left half, so the minus sign

Find the inverse function of f(x) = x2 − 4 on the restricted domain x ≤ 0, and state its domain.

Show the steps and the answer
(−4, 0)(−3, −1)(5, −3)domainrange−448−4−22
The graph of −x+4 starts at x = −4 and never rises above 0, so every output lands in the kept half x ≤ 0.
  1. Say what is asked: f squares an input that is 0 or less, then subtracts 4. Find the rule that returns that input, which is 0 or negative.The answers of the inverse must land back in x ≤ 0, the half that was kept.
  2. Write y=x2−4, with the note x≤0. Swap: x=y2−4, with the note y≤0.The letters trade places, and the note now speaks about y.
  3. Add 4 to both sides: x+4=y2.Subtracting 4 was the last step of f, so adding 4 undoes it and leaves the square alone.
  4. Take the square root of both sides: y=±x+4.Both signs are candidates until the note is read.
  5. Choose the sign. The note says y≤0, so keep the minus sign: y=−x+4. Rename: f−1(x)=−x+4.x+4 is 0 or more, so its negative is 0 or less for every x in the domain. The plus root is above 0 for every x except −4.
  6. State the domain: x≥−4.x2 is 0 or more, so the outputs of f are −4 or more. The root agrees: x + 4 must be 0 or more.
Answer
  • f⁻¹(x)=−x+4
  • Domain of f⁻¹: x ≥ −4, which is [−4, ∞)
Double-check f(−3) = 9 − 4 = 5, then f⁻¹(5) = −5+4 = −9 = −3. The plus root would give 3, which is not in x ≤ 0. Again with −1: f(−1) = 1 − 4 = −3, then f⁻¹(−3) = −−3+4 = −1 = −1.

Rung 4 of 6

A shifted vertex and the left half

Find the inverse function of f(x) = (x + 1)2 − 6 on the restricted domain x ≤ −1, and state its domain.

Show the steps and the answer
−4f: add 1,square,subtract 6f⁻¹: add 6,take the root,make itnegative,firstsecond3−4
The input −4 becomes 3 under f and returns to −4 under f⁻¹, with the minus sign doing the work of the left…
  1. Say what is asked: f takes an input that is −1 or less, adds 1, squares, and subtracts 6. Find the rule that goes back to that input.The bracket x + 1 is zero at x = −1, so the vertex is there and the left half was kept.
  2. Write y=(x+1)2−6, with the note x≤−1. Swap: x=(y+1)2−6, with the note y≤−1.Input and output trade letters, and the note goes with the old inputs.
  3. Add 6 to both sides: x+6=(y+1)2.Subtracting 6 was the last step of f, so it is undone first.
  4. Take the square root of both sides: y+1=±x+6.The whole bracket y + 1 was squared, so the root gives back the whole bracket, with both signs possible.
  5. Choose the sign. The note says y ≤ −1, so y + 1 is 0 or less. Keep the minus sign: y+1=−x+6.The minus root is 0 or less for every x ≥ −6. The plus root is above 0 for every x except −6.
  6. Subtract 1 from both sides: y=−1−x+6. Rename: f−1(x)=−1−x+6.This undoes the add 1 inside the bracket, the first thing f did and the last thing to undo.
  7. State the domain: x≥−6.(x + 1)2 is 0 or more, so the outputs of f are −6 or more. The root agrees: x + 6 must be 0 or more.
Answer
  • f⁻¹(x)=−1−x+6
  • Domain of f⁻¹: x ≥ −6, which is [−6, ∞)
Double-check f(−4) = (−4 + 1)2 − 6 = (−3)2 − 6 = 9 − 6 = 3, then f⁻¹(3) = −1 − 3+6 = −1 − 3 = −4. Again with −2: f(−2) = (−1)2 − 6 = −5, then f⁻¹(−5) = −1 − −5+6 = −1 − 1 = −2.

Rung 5 of 6

A number multiplying the square

Find the inverse function of f(x) = 3x2 − 2 on the restricted domain x ≥ 0, and state its domain.

Show the steps and the answer
=x + 23y²do the same thing to both sides
Both sides must be divided by 3 as whole sides.
  1. Say what is asked: f squares an input that is 0 or more, multiplies by 3, then subtracts 2. Find the rule that goes back.Three steps were done, so three steps are undone, in backward order.
  2. Write y=3x2−2, with the note x≥0. Swap: x=3y2−2, with the note y≥0.Input and output trade letters, and the note follows.
  3. Add 2 to both sides: x+2=3y2.Subtracting 2 was the last step of f, so it is undone first.
  4. Divide both sides by 3: x+23=y2.Multiplying by 3 was the step before that. The whole left side is divided, not only the x. The square now stands alone.
  5. Take the square root of both sides: y=±(x+2)÷3.The root covers the whole fraction. (x + 2) ÷ 3 is the same fraction written on one line so that all of it sits under the root.
  6. Choose the sign. The note says y≥0, so y=(x+2)÷3. Rename: f−1(x)=(x+2)÷3.The plus root is 0 or more, as the note requires.
  7. State the domain: x≥−2.3x2 is 0 or more, so the outputs of f are −2 or more. The root agrees: x + 2 must be 0 or more.
Answer
  • f⁻¹(x)=(x+2)÷3
  • Domain of f⁻¹: x ≥ −2, which is [−2, ∞)
Double-check f(2) = 3 · 4 − 2 = 10, then f⁻¹(10) = (10+2)÷3 = 12÷3 = 4 = 2. Again with 4: f(4) = 3 · 16 − 2 = 46, then f⁻¹(46) = 48÷3 = 16 = 4. Gotcha: dividing first and adding 2 afterward gives 10÷3+2, which is about 2.31 (rounded), not 2.

Rung 6 of 6

A parabola that opens down, left half kept

Find the inverse function of f(x) = −2(x − 1)2 + 8 on the restricted domain x ≤ 1, and state its domain.

Show the steps and the answer
8(−∞, 8]
This parabola opens down with its top at 8, so the inverse accepts only inputs of 8 or less.
  1. Say what is asked: f takes an input that is 1 or less, subtracts 1, squares, multiplies by −2, and adds 8. Find the rule that goes back to that input.The bracket x − 1 is zero at x = 1, so the vertex is there and the left half was kept.
  2. Write y=−2(x−1)2+8, with the note x≤1. Swap: x=−2(y−1)2+8, with the note y≤1.Input and output trade letters, and the note goes with the old inputs.
  3. Subtract 8 from both sides: x−8=−2(y−1)2.Adding 8 was the last step of f, so it is undone first.
  4. Divide both sides by −2: x−8−2=(y−1)2. Multiply the top and the bottom of the fraction by −1 to tidy it: 8−x2=(y−1)2.Dividing undoes the multiplying by −2 and leaves the squared bracket alone. Multiplying top and bottom by −1 does not change a fraction: with x = 0, both versions equal 4.
  5. Take the square root of both sides: y−1=±(8−x)÷2.The root covers the whole fraction, and both signs are candidates for now.
  6. Choose the sign. The note says y ≤ 1, so y − 1 is 0 or less. Keep the minus sign: y−1=−(8−x)÷2.The minus root is 0 or less for every x ≤ 8. The plus root is above 0 for every x except 8.
  7. Add 1 to both sides: y=1−(8−x)÷2. Rename: f−1(x)=1−(8−x)÷2.This undoes the subtract 1 inside the bracket, the first step of f and the last to undo.
  8. State the domain: x≤8.(x − 1)2 is 0 or more, so −2(x − 1)2 is 0 or less, and the outputs of f are 8 or less. The root agrees: 8 − x ≥ 0 gives −x ≥ −8, and dividing by −1 flips the sign to x ≤ 8.
Answer
  • f⁻¹(x)=1−(8−x)÷2
  • Domain of f⁻¹: x ≤ 8, which is (−∞, 8]
Double-check f(−1) = −2(−1 − 1)2 + 8 = −2 · 4 + 8 = 0, then f⁻¹(0) = 1 − (8−0)÷2 = 1 − 4 = −1. Again with −2: f(−2) = −2(−3)2 + 8 = −18 + 8 = −10, then f⁻¹(−10) = 1 − (8+10)÷2 = 1 − 9 = −2.

Counterexamples

What it is not, and the tempting wrong moves.

✗ Not this: For f(x) = x2 − 4 on x ≤ 0, writing f⁻¹(x) = x+4.
The symbol √ gives 0 or more, but every answer must land in x ≤ 0. Test it: f(−3) = 5, yet 5+4 = 3, not −3.
✓ Instead: The left half was kept, so take the minus sign: f⁻¹(x) = −x+4. Then f⁻¹(5) = −3.
✗ Not this: Leaving both signs in the answer: f⁻¹(x) = 3 ± x−2.
That gives two outputs for one input, for example 5 and 1 when x = 6. A function gives one output for each input, so this is not a function.
✓ Instead: Use the restricted domain to keep one sign. For x ≥ 3 the answer is f⁻¹(x) = 3 + x−2.
✗ Not this: Swapping x and y in the equation but leaving the note as x ≥ 3.
After the swap the old inputs are called y. A note that still says x ≥ 3 now describes the wrong letter, and it gives no help when the sign is chosen.
✓ Instead: Swap the note too: x ≥ 3 becomes y ≥ 3. Then y − 3 is 0 or more, and the plus sign follows.
✗ Not this: Taking the root too early: from x = y2 + 5, writing x = y + 5.
A root cannot be split across a plus sign. 9+16 = 25 = 5, but 9 + 16 = 3 + 4 = 7.
✓ Instead: Move the 5 first: x − 5 = y2. Then take the root of each whole side: y = ±x−5.
✗ Not this: Reading f⁻¹(x) as 1 ÷ f(x).
The small −1 is part of the name, not a power. For f(x) = x2 + 5 on x ≥ 0, f(2) = 9, so the inverse must send 9 back to 2. But 1 ÷ f(9) = 186.
✓ Instead: f⁻¹(x) = x−5, and f⁻¹(9) = 4 = 2.
✗ Not this: Looking for the inverse function of f(x) = x2 − 4 on all real numbers, with no restricted domain.
The method does not apply here. f(3) = 5 and f(−3) = 5, so the output 5 has two inputs and f is not a one-to-one function. No single sign can send 5 back to both.
✓ Instead: Cut the domain at the vertex first. On x ≥ 0 the inverse is x+4. On x ≤ 0 it is −x+4.

Tips and tricks

  • Memory device: right half, plus root. Left half, minus root.
  • Check with a number that is not the vertex. At the vertex the root is 0, so both signs give the same value and a wrong sign slips through.
  • Read f as a list of steps, then write the list backward with opposite steps. For (x − 3)2 + 2: subtract 3, square, add 2. Backward: subtract 2, take the root, add 3. That is 3 + x−2.
  • Domain and range trade places. Whatever comes out of f goes into f⁻¹, and whatever went into f comes out of f⁻¹.
  • Do not memorize the general formula in the rule. The steps rebuild it in ten seconds, and they still work when the numbers are awkward.
  • In the finished answer, the number standing outside the root is the x value of the vertex, the place where the domain was cut. If it is not, recheck the last step.

7 The trap

Writing the plus root out of habit. The root step always produces ±, and the plus sign is right only when the right half of the parabola was kept. When the restricted domain is x ≤ h, the answer needs the minus sign. To avoid the mistake, carry the domain note through the swap, read it before you choose the sign, and test one number that is not the vertex.

Flash cards

Say the answer out loud, then flip.

The square root step leaves you with ±. How do you choose the sign?
  • Read the domain note after the swap.
  • If it says y ≥ h, keep the plus sign.
  • If it says y ≤ h, keep the minus sign.
Why does undoing a square give two candidates?
  • Two numbers share each square above 0: 32 = 9 and (−3)2 = 9. Only 0 comes from one number, 0.
  • It is like a friend saying "I am 2 blocks from your home" without saying east or west.
  • The restricted domain is the missing direction, so it picks the sign.
Find the inverse function of f(x) = (x − 2)2 + 1 on the restricted domain x ≤ 2.
  • f⁻¹(x) = 2 − x−1, with domain x ≥ 1.
  • The sign is minus because the left half of the parabola was kept.
  • Check with 0: f(0) = 5, and f⁻¹(5) = 2 − 4 = 0.
Once you have f⁻¹, how do you find its domain without new work?
  • It is the range of f.
  • The outputs of f are the inputs of f⁻¹.
  • In the same way, the range of f⁻¹ is the restricted domain of f.

Lesson 3 of 6

Inverses of other polynomials

Subject: MathematicsLeans on: Cubing a numberLeans on: The horizontal line testLeans on: Swap and solveLeans on: Keeping an equation balancedLeans on: A negative number in the bottom of a fractionLeans on: Reading (−∞, ∞)Open this lesson
Leans on 0Comes back in 0

1 In plain words

A polynomial is a sum of terms such as 2x3 or −5, where x carries whole-number powers. A cubic is a polynomial whose highest power is x3, and x3 means x · x · x. Picture a coat check. Squaring is a careless coat check: 3 and −3 both get ticket 9, so the ticket cannot say whose coat it is. Cubing is a careful one: 3 gets ticket 27 and −3 gets ticket −27. Every ticket points back to one person. So a cubic such as x3 + 4 has an inverse function on its whole domain, and nothing has to be cut. Reading a ticket back is the cube root: 273 = 3 because 3 · 3 · 3 = 27.

2 Picture it

(2, 8)(1, 1)(−2, −8)−22−8−448(2, 8)(1, 1)(−2, −8)−2−112−8−448
Each horizontal line meets the graph of y = x3 at one point, so every output comes from one input.

Reminder

Reminder: Cubing a number, The horizontal line test, Swap and solve, Keeping an equation balanced, A negative number in the bottom of a fraction, Reading (−∞, ∞)
  • Cubing a number. x3 means x · x · x. So 23 = 2 · 2 · 2 = 8, and (−2)3 = (−2) · (−2) · (−2) = 4 · (−2) = −8. A negative cubed stays negative.
  • The horizontal line test. If no horizontal line meets a graph twice, every output comes from one input, and the function is a one-to-one function. y = x2 fails at height 4, where x = −2 and x = 2 both land.
  • Swap and solve. To find an inverse, write y for f(x), swap x and y, then solve for y. For y = x + 5: swap to x = y + 5, then y = x − 5.
  • Keeping an equation balanced. Whatever is done to one side is done to the other. From x − 6 = 2y3, dividing both sides by 2 gives x−62 = y3. The whole left side is divided, not only the x.
  • A negative number in the bottom of a fraction. Multiplying the top and the bottom by −1 keeps the value and moves the sign: x−1−2 = 1−x2. With x = 5, both sides are −2.
  • Reading (−∞, ∞). The interval (−∞, ∞) means every real number, with no number left out.

3 Why

Cubing keeps the sign of a number: a positive cubed is positive, and a negative cubed is negative. Cubing also keeps the order: a larger number always has a larger cube. So two different inputs can never share an output, and every horizontal line meets the graph of y = x3 exactly once. Shifting, stretching or flipping that graph does not add a second meeting point. Each output has one input, so the cube root gives one answer and no ± choice is needed. Squaring throws the sign away, and that is the only reason a quadratic needed a restricted domain.

4 The rule

A cubic of the form f(x)=a(x−h)3+k, with a≠0, is a one-to-one function on all real numbers, so it has an inverse function with no restricted domain.

To find it: write y=f(x), swap x and y, and solve for y, using a cube root to undo the cube. A cube root has one value and takes no ± sign. The result is f−1(x)=(x−ka)3+h, with domain (−∞,∞) and range (−∞,∞).

The same idea, five ways

Say it

f inverse of x equals the cube root of x. Say: the cube root of 8 is 2.

Write it

The inverse function of the cubing function is the cube root function: it takes a number and returns the one number whose cube is that number.

In math
  • f(x) = x3 and f−1(x) = x3
  • y = x3 means the same as x = y3
  • 83 = 2 because 23 = 8
  • ∛(−8) = −2 because (−2)3 = −8
  • Domain of f−1: (−∞, ∞). Range of f−1: (−∞, ∞).
Like

A coat check that gives every coat its own ticket, so any ticket can be traded back for exactly one coat.

See it
3cube it: x³cube root: ∛xfirstsecond273
The cube root machine reads the output 27 back to the single input 3.

The same idea, other ways

As a picture

Draw y = x3 and, dashed behind it, y = x2. Lay a ruler flat at height 4. The ruler meets the dashed parabola twice, at x = −2 and x = 2, so the output 4 has two inputs. The same ruler meets the cubic once, near x = 1.59. Slide the ruler to any height, even below the x-axis: the cubic is always met exactly once. One meeting point per height means the function can be run backward without cutting anything.

(−2, 4)(2, 4)(∛4, 4)−22−8−448
The line at height 4 meets the dashed parabola twice and the solid cubic once.
With small numbers

Look at the table for x3. The inputs −2, −1, 0, 1, 2 give the outputs −8, −1, 0, 1, 8. Find 8 in the output row: it appears once, under the input 2. Find −8: it appears once, under −2. No output repeats. Now imagine the same table for x2. The outputs would be 4, 1, 0, 1, 4, and both 4 and 1 would appear twice. A repeated output is what forces a cut. The cubic has none.

input xoutput x³−2−8−1−1001128↑ solve: output given, read every input above it
The output 8 sits in one cell only, under the input 2, so 83 has one answer.
As a story

Putting on socks and then shoes is undone by taking off shoes and then socks: last on, first off. The function f(x) = 2x3 + 6 does three things in order: cube, double, add 6. Its inverse takes them off in reverse order: subtract 6, halve, cube root. Because each of those three steps can be undone in one way only, the whole function can be undone in one way only.

Rule of thumb

Odd keeps, even loses. An odd power such as x3 keeps the sign of the input, so nothing is lost and the function can be reversed everywhere. An even power such as x2 loses the sign, so half of the inputs must be cut away before it can be reversed.

FunctionDo two inputs share an output?Cut needed?Undo step
f(x)=x2Yes. (−3)2 = 9 and 32 = 9.Yes, at the vertex.Square root, with the sign chosen to match the cut.
f(x)=x3No. (−3)3 = −27 and 33 = 27.No.Cube root. One answer, no ±.
f(x)=2(x+1)3−3No. It is x3 shifted and stretched.No.Add 3, divide by 2, cube root, subtract 1.
f(x)=x3−xYes. The inputs −1, 0 and 1 all give 0.It has no inverse on all real numbers.The method of this lesson does not apply.

5 How

  1. Look at the shape. If x appears once, inside a single cube, the function is x3 shifted, stretched or flipped. It is one-to-one and needs no cut.
  2. Write y in place of f(x).
  3. Swap x and y. The equation now describes the inverse.
  4. Undo whatever sits outside the cube, last thing done first: move an added or subtracted number to the other side, then divide or multiply away the number in front.
  5. With the cube alone on its side, take the cube root of both sides. Write no ± sign.
  6. If anything is still attached to y, such as y − 2, undo it. That number lands outside the root.
  7. Rename y as f−1(x). State the domain and the range: both are all real numbers.
  8. Check with a number. Put an input into f, feed the output into f−1, and confirm that the input comes back.

Finding the inverse of a cubic with one cube

1
Does x appear once, inside a single cube, as in a(x − h)3 + k?
YesThe function is one-to-one. No cut is needed. Write y for f(x), swap x and y, and go on.
NoTest a few inputs for a repeated output, as with x3 − x at −1, 0 and 1. If an output repeats, there is no inverse on all real numbers and this method stops.
↓
2
After the swap, is a number added to or subtracted from the cube term?
YesMove it to the other side by doing the opposite to both sides.
NoGo on.
↓
3
Is the cube multiplied or divided by a number?
YesDivide or multiply both sides by that number, so the cube stands alone.
NoGo on.
↓
4
Is the cube alone on its side?
YesTake the cube root of both sides. One value, no ± sign.
NoGo back and finish clearing the outside of the cube.
↓
5
Is anything still attached to y, such as y − 2 or y + 1?
YesUndo it on both sides. The number ends up outside the root.
NoRename y as f−1(x), state the domain and range, and check with a number.
  1. Check that x appears once, inside a single cube. If so, the function is one-to-one and needs no cut.
  2. Write y for f(x), then swap x and y.
  3. Clear the side holding y until the cube is alone: first move any added or subtracted number, then divide or multiply away the number attached to the cube.
  4. Take the cube root of both sides. Write no ± sign.
  5. Undo anything still attached to y. That number goes outside the root.
  6. Rename y as f−1(x) and state the domain and range, both (−∞, ∞).
  7. Check with a round trip: one positive input and one negative input should each come back.

6 Worked example

Worked example

The inverse of a stretched and raised cubic

Find the inverse function of f(x) = 2x3 + 6, and state its domain and range.

Show the steps and the answer
input xoutput f(x) = 2x³ + 6−2−10−140618222↓ evaluate: input given, read the output below it
Every output in the bottom row appears once, so each one can be traced back to a single input.
2f: cube,double, add 6f inverse:subtract 6,halve, cuberootfirstsecond222
The second machine reverses the first, step by step, so the input 2 comes back out.
  1. Say what the question asks: build the rule that takes each output of f back to the one input it came from.An inverse function runs the original function backward. If f turns 2 into 22, the inverse must turn 22 into 2.
  2. Check the shape: x appears once, inside one cube. This is x3 stretched by 2 and raised by 6.A stretched and raised copy of x3 still meets every horizontal line once, so f is a one-to-one function and no restricted domain is needed.
  3. Write y=2x3+6.y is the name of the output. Naming it makes the swap in the next step visible.
  4. Swap x and y: x=2y3+6.The inverse trades the roles of input and output, so the two letters trade places.
  5. Subtract 6 from both sides: x−6=2y3.f cubes first, then doubles, then adds 6. Adding 6 was done last, so it is undone first. This step clears everything from the right side except the term holding y.
  6. Divide both sides by 2: x−62=y3.Doubling is the next thing to undo. This step leaves the cube alone on its side, ready for the cube root.
  7. Take the cube root of both sides: y=(x−62)3.The cube root undoes the cube. Only one number has a given cube, so there is one answer and no ± sign.
  8. Rename y: f−1(x)=(x−62)3. Domain: (−∞,∞). Range: (−∞,∞).f accepts every real number and its outputs cover every real number. The inverse's inputs are f's outputs, and the inverse's outputs are f's inputs, so both lists are all real numbers.
Answer
  • f−1(x)=∛(x−62)
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check A round trip with numbers. In the table, the column under 2 shows f(2) = 2 · 8 + 6 = 22. Then f−1(22) = ∛(22−62) = ∛(162) = 83 = 2, the starting input. A negative input works too: f(−1) = 2 · (−1) + 6 = 4, and f−1(4) = ∛(4−62) = ∛(−1) = −1. A second way to the formula: f cubes, doubles, then adds 6, so the reverse is subtract 6, halve, then cube root, which is the same rule.

Ladder: from easy to exam-hard

Try each rung before you open its steps.

Rung 1 of 7

Rung 1: the plain cube

Find the inverse function of f(x) = x3.

Show the steps and the answer
−2cube it: x³cube root: ∛xfirstsecond−8−2
Cubing and then taking the cube root returns the starting number, negative sign included.
  1. Say what the question asks: find the rule that takes each cube back to the number that was cubed.The inverse must turn 27 back into 3 and −8 back into −2.
  2. Write y=x3.y names the output.
  3. Swap x and y: x=y3.The inverse trades input and output.
  4. Take the cube root of both sides: x3=y.The cube root undoes the cube and has one value, so no ± sign appears.
  5. Rename y: f−1(x)=x3.y now stands for the output of the inverse.
Answer
  • f−1(x)=x3
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check f(3) = 27 and f−1(27) = 273 = 3. With a negative: f(−2) = −8 and f−1(−8) = ∛(−8) = −2. Both inputs come back.

Rung 2 of 7

Rung 2: a number added outside the cube

Find the inverse function of f(x) = x3 + 4.

Show the steps and the answer
  1. Write y = x3 + 4, then swap x and y: x=y3+4.Swapping the letters turns the equation for f into the equation for its inverse.
  2. Subtract 4 from both sides: x−4=y3.Adding 4 was the last thing f did, so it is the first thing to undo. This leaves the cube alone.
  3. Take the cube root of both sides: y=(x−4)3.The cube is alone, so the cube root removes it. One value, no ±.
  4. Rename y: f−1(x)=(x−4)3.This is the finished rule of the inverse.
Answer
  • f−1(x)=∛(x−4)
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check f(2) = 8 + 4 = 12, and f−1(12) = ∛(12 − 4) = 83 = 2. Also f(−1) = −1 + 4 = 3, and f−1(3) = ∛(3 − 4) = ∛(−1) = −1.

Rung 3 of 7

Rung 3: a shift inside the cube

Find the inverse function of f(x) = (x − 2)3.

Show the steps and the answer
5subtract 2,then cubecube root,then add 2firstsecond275
The inverse adds 2 after the cube root, because f subtracted 2 before the cube.
  1. Write y = (x − 2)3, then swap x and y: x=(y−2)3.The swap describes the inverse.
  2. Take the cube root of both sides right away: x3=y−2.Nothing sits outside the cube, so the cube itself is the last thing done and the first thing undone.
  3. Add 2 to both sides: y=x3+2.Subtracting 2 happened inside the cube, before cubing, so it is undone after the cube root. The + 2 lands outside the root.
  4. Rename y: f−1(x)=x3+2.Here x3 + 2 means the cube root of x, and then 2 added to the result.
Answer
  • f−1(x) = x3 + 2, with the 2 outside the root
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check f(5) = (5 − 2)3 = 33 = 27, and f−1(27) = 273 + 2 = 3 + 2 = 5. Also f(0) = (−2)3 = −8, and f−1(−8) = ∛(−8) + 2 = −2 + 2 = 0.

Rung 4 of 7

Rung 4: a negative number in front

Find the inverse function of f(x) = 1 − 2x3.

Show the steps and the answer
  1. Write y = 1 − 2x3, then swap x and y: x=1−2y3.The swap describes the inverse.
  2. Subtract 1 from both sides: x−1=−2y3.The 1 is not attached to the cube, so it moves first. The minus sign stays with the 2.
  3. Divide both sides by −2: x−1−2=y3.The cube is multiplied by −2, so dividing by −2 leaves the cube alone.
  4. Tidy the sign: x−1−2=1−x2, so 1−x2=y3.Multiplying the top and the bottom of a fraction by −1 does not change its value: the top x − 1 becomes 1 − x and the bottom −2 becomes 2.
  5. Take the cube root of both sides and rename: f−1(x)=(1−x2)3.The cube is alone, and the cube root has one value.
Answer
  • f−1(x)=∛(1−x2)
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check f(2) = 1 − 2 · 8 = 1 − 16 = −15, and f−1(−15) = ∛(1−(−15)2) = ∛(162) = 83 = 2. Also f(−1) = 1 − 2 · (−1) = 3, and f−1(3) = ∛(1−32) = ∛(−1) = −1.

Rung 5 of 7

Rung 5: a stretch, an inside shift and an outside shift together

Find the inverse function of f(x) = 2(x + 1)3 − 3.

Show the steps and the answer
  1. Write y = 2(x + 1)3 − 3, then swap x and y: x=2(y+1)3−3.The swap describes the inverse.
  2. Add 3 to both sides: x+3=2(y+1)3.f adds 1, cubes, doubles, then subtracts 3. Subtracting 3 was last, so it is undone first.
  3. Divide both sides by 2: x+32=(y+1)3.Doubling is the next step to undo. Now the cube is alone.
  4. Take the cube root of both sides: (x+32)3=y+1.The cube root removes the cube and gives one value.
  5. Subtract 1 from both sides and rename: f−1(x)=(x+32)3−1.Adding 1 was the first thing f did, so it is the last thing undone. The − 1 sits outside the root.
Answer
  • f−1(x)=∛(x+32)−1
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check f(1) = 2 · 23 − 3 = 16 − 3 = 13, and f−1(13) = ∛(13+32) − 1 = 83 − 1 = 2 − 1 = 1. Also f(−2) = 2 · (−1)3 − 3 = −2 − 3 = −5, and f−1(−5) = ∛(−5+32) − 1 = ∛(−1) − 1 = −1 − 1 = −2.

Rung 6 of 7

Rung 6: the cube divided by a number

Find the inverse function of f(x) = (x−3)34 + 2.

Show the steps and the answer
  1. Write y=(x−3)34+2, then swap x and y: x=(y−3)34+2.The swap describes the inverse.
  2. Subtract 2 from both sides: x−2=(y−3)34.Adding 2 was the last thing f did, so it goes first.
  3. Multiply both sides by 4: 4(x−2)=(y−3)3, which is 4x−8=(y−3)3.The cube is divided by 4, and multiplying by 4 undoes that. The 4 multiplies the whole left side, so both x and −2 are multiplied.
  4. Take the cube root of both sides: (4x−8)3=y−3.The cube is alone, and the cube root has one value.
  5. Add 3 to both sides and rename: f−1(x)=(4x−8)3+3.Subtracting 3 was the first thing f did, so it is undone last. The + 3 sits outside the root.
Answer
  • f−1(x)=∛(4x−8)+3
  • Domain: (−∞, ∞)
  • Range: (−∞, ∞)
Double-check f(5) = 234 + 2 = 84 + 2 = 2 + 2 = 4, and f−1(4) = ∛(16 − 8) + 3 = 83 + 3 = 2 + 3 = 5. Also f(1) = (−2)34 + 2 = −84 + 2 = −2 + 2 = 0, and f−1(0) = ∛(0 − 8) + 3 = −2 + 3 = 1.

Rung 7 of 7

Rung 7: a cubic that this method cannot invert

Does f(x) = x3 − x have an inverse function on all real numbers?

Show the steps and the answer
input xoutput x³ − x−2−6−10001026↑ solve: output given, read every input above it
The output 0 appears under three inputs, so this cubic fails the horizontal line test.
  1. Say what the question asks: decide whether every output of f comes from one input only.Only a one-to-one function has an inverse function.
  2. Notice the shape: x appears in two places, in x3 and in −x.The no-cut promise covers a cubic with x inside one cube. This one is not of that shape, so it must be tested.
  3. Compute three outputs. f(−1)=(−1)3−(−1)=−1+1=0. f(0)=0−0=0. f(1)=1−1=0.Small inputs are quick to test, and the table shows the result: the output 0 sits under three different inputs.
  4. Conclude: the horizontal line at height 0 meets the graph three times, so f is not a one-to-one function.An inverse would have to send 0 back to −1, 0 and 1 at once, and a function gives one output per input.
Answer
No. The inputs −1, 0 and 1 all give the output 0, so f has no inverse function on all real numbers.
Double-check A second way, by factoring: x3 − x = x(x − 1)(x + 1). Multiplying back gives x(x2 − 1) = x3 − x, so the factoring is right. A product is 0 when any one bracket is 0, which happens at x = 0, x = 1 and x = −1: the same three inputs.

Counterexamples

What it is not, and the tempting wrong moves.

✗ Not this: Solving y3 = x − 4 as y = ±∛(x − 4), the way a square is undone.
The ± belongs to squares, where two numbers share a square: 32 = 9 and (−3)2 = 9. Cubes are not shared: 23 = 8 but (−2)3 = −8. Only one number cubes to x − 4.
✓ Instead: y = ∛(x − 4), one value, no ± sign.
✗ Not this: Cutting the domain of f(x) = x3 + 4 to x ≥ 0 before inverting, the way a parabola is cut at its vertex.
The cut throws away half of a function that was already one-to-one. With the cut, f(−2) = −4 would be lost, yet the inverse handles it without trouble: ∛(−4 − 4) = ∛(−8) = −2.
✓ Instead: Keep every real number. The domain of f and the domain of f−1 are both (−∞, ∞).
✗ Not this: Saying ∛(−8) is not a real number, because a root cannot hold a negative.
That limit belongs to square roots, since no real number squared is negative. A negative number cubed is negative: (−2) · (−2) · (−2) = −8.
✓ Instead: ∛(−8) = −2. A cube root accepts any real number.
✗ Not this: For f(x) = (x − 2)3, writing f−1(x) = ∛(x + 2) with the 2 inside the root.
Test it. f(5) = 27, but ∛(27 + 2) = 293, which is about 3.07 and not 5. The shift happened inside the cube, so it is undone after the cube root, outside it.
✓ Instead: f−1(x) = x3 + 2, and then 273 + 2 = 3 + 2 = 5.
✗ Not this: Believing every cubic is one-to-one, and trying to invert f(x) = x3 − x by swap and solve.
f(−1), f(0) and f(1) are all 0, so the output 0 has three inputs. The rule covers a cubic where x sits inside one cube, such as a(x − h)3 + k.
✓ Instead: When x appears in more than one term, test for a repeated output first. If one is found, the function has no inverse on all real numbers.

Tips and tricks

  • Memory device: odd keeps, even loses. An odd power keeps the sign, so x3 needs no cut: (−2)3 = −8 and 23 = 8. An even power loses it, so x2 needs a cut: (−2)2 = 4 and 22 = 4. The same holds for higher powers: (−2)5 = −32 and 25 = 32 differ, while (−2)4 and 24 are both 16.
  • Do not memorize the formula f−1(x) = ∛(x−ka) + h. Swap and solve rebuilds it in ten seconds, and the rebuild is harder to get wrong than a remembered formula.
  • Undo in reverse order: last on, first off, like shoes and socks. List what f does to x in order, then reverse the list and reverse each step.
  • Know the small cubes cold, since they make checking fast: 13 = 1, 23 = 8, 33 = 27, 43 = 64, 53 = 125. Their negatives give −1, −8, −27, −64, −125.
  • For the check, pick an input that makes the arithmetic clean, then test one negative input as well. A sign mistake often hides until a negative number is tried.
  • A number that sat outside the cube in f ends up inside the root in f−1. A number that sat inside the cube in f ends up outside the root in f−1.

7 The trap

The most common mistake is carrying square root habits into cube roots: writing ± in front of the cube root, or cutting the domain as if the cubic were a parabola. A cube root has one value. 83 is 2 only, because (−2)3 is −8 and not 8. Before writing ±, ask whether the power being undone is even or odd. Only an even power loses the sign.

Flash cards

Say the answer out loud, then flip.

Why does a cubic such as x3 + 4 need no cut in its domain before you find its inverse?
Cubing keeps order: the larger of two numbers always has the larger cube, so two different inputs never share an output. It is a careful coat check: 3 gets ticket 27 and −3 gets ticket −27, so every ticket points back to one person.
When you undo a cube with a cube root, do you write ± in front of the root?
No. A cube root has one value. 83 is 2 only, because (−2)3 = −8 and not 8. Only an even power, such as a square, loses the sign.
Find the inverse of f(x) = 2(x − 1)3 + 5.
  • Swap x and y: x = 2(y − 1)3 + 5.
  • Subtract 5, then divide by 2: x−52 = (y − 1)3.
  • Take the cube root, then add 1: f−1(x) = ∛(x−52) + 1.
  • Check: f(3) = 2 · 8 + 5 = 21, and f−1(21) = 83 + 1 = 3.
What are the domain and the range of the inverse of f(x) = a(x − h)3 + k, with a ≠ 0?
  • Domain: all real numbers, (−∞, ∞).
  • Range: all real numbers, (−∞, ∞).
  • Nothing was cut from the cubic, so nothing is missing from its inverse.

Lesson 4 of 6

Radical functions and their domains

Subject: MathematicsLeans on: What a domain isLeans on: Solving a linear inequalityLeans on: Flipping an inequalityLeans on: Interval notationLeans on: Signs of squares and cubesLeans on: Dividing signed numbersOpen this lesson
Leans on 1Comes back in 0

1 In plain words

A radical function is a rule with x under a root sign, such as 2x+6 or x3. Picture the root sign as a gate with a guard. The square root guard turns away every negative number, because no real number times itself gives a negative. So the expression under the root, called the inside, must come out 0 or more. The domain is the set of x values that get past the guard. A square root function always has this guard. A cube root function has an open gate: it takes negatives, zero and positives alike, so its domain is every real number.

2 Picture it

(−3, 0)(−1, 2)(5, 4)domainrange−448−224(−3, 0)(−1, 2)(5, 4)domainrange−6−4−22468−224
The graph of y = 2x+6 starts at (−3, 0) and runs right, so its shadow on the x-axis, the domain.

Reminder

Reminder: What a domain is, Solving a linear inequality, Flipping an inequality, Interval notation, Signs of squares and cubes, Dividing signed numbers
  • What a domain is. The domain is the set of inputs a function accepts. For x the input 9 is in the domain (9 = 3) and the input −9 is not.
  • Solving a linear inequality. Solve it like an equation, one undo at a time: 2x + 6 ≥ 0 becomes 2x ≥ −6, then x ≥ −3.
  • Flipping an inequality. Multiplying or dividing both sides by a negative flips the sign: −2x ≥ −10 becomes x ≤ 5.
  • Interval notation. A square bracket includes the endpoint and a round bracket leaves it out. x ≥ −3 is [−3, ∞) and x ≤ 5 is (−∞, 5]. Infinity always gets a round bracket.
  • Signs of squares and cubes. Two negatives multiply to a positive: (−3)2 = 9. Three negatives multiply to a negative: (−2)3 = −8.
  • Dividing signed numbers. A negative divided by a negative is positive: −10 ÷ (−2) = 5. A negative divided by a positive is negative: −6 ÷ 2 = −3.

3 Why

Squaring never gives a negative: 32 = 9, (−3)2 = 9 and 02 = 0. So no real number squares to −9, and −9 is not a real number. That is the whole reason the inside of a square root must be 0 or more. Cubing is different because it keeps the sign: 23 = 8 and (−2)3 = −8. So ∛(−8) = −2 is a real answer. Every real number, negative or not, is the cube of exactly one real number, which is why a cube root never turns an input away.

4 The rule

Square root (and every even root): f(x)=ax+b gives a real number only when ax+b≥0.

Solve that inequality. Its solution is the domain.
Cube root (and every odd root): g(x)=(ax+b)3 gives a real number for every x. Its domain is (−∞,∞).

The same idea, five ways

Say it

f of x equals the square root of 2x plus 6, and its domain is x at least negative 3. For a cube root say: the domain is all real numbers.

Write it

A radical function is a function with the variable under a root sign. A square root function needs its inside to be greater than or equal to 0. A cube root function accepts every real number.

In math
  • f(x) = ax+b needs ax + b ≥ 0
  • f(x) = 2x+6 has domain x ≥ −3
  • [−3,∞)
  • {x|x≥−3}
  • g(x) = ∛(ax + b) has domain (−∞, ∞)
  • Graph words: the square root curve starts at x = −3 and runs to the right; the cube root curve runs left and right forever
Like

A square root is a gate with a guard who turns away anything below 0. A cube root is the same gate with no guard.

See it
−3[−3, ∞)
The filled dot at −3 says −3 is allowed, and the shading to the right is every other allowed input of 2x+6.

The same idea, other ways

As a picture

Look at the graph of y = x2. It sits on or above the x-axis everywhere. Asking for −9 is asking where that curve reaches the height −9. It never does. Asking for 9 is asking where it reaches the height 9. It does, at x = −3 and at x = 3, and the root symbol takes the one that is 0 or more, so 9 = 3. A square root can only be asked about heights the curve reaches: 0 and up.

no square ever lands below 0(−3, 9)(3, 9)(0, 0)−33−49
The curve y = x2 never goes below the x-axis, so no real number has a negative square.
As a story

You are laying a square patio. If the area is 25 square feet, each side is 25 = 5 feet. If the area is 0, the side is 0: no patio, but still an answer. An area of −9 square feet makes no sense, so there is no side length to report. In f(x) = 2x+6, the inside 2x + 6 plays the part of the area. You may only choose x values that make that area 0 or more.

With small numbers

Read the table for the inside, 2x + 6. Under x = −5 and x = −4 the inside is negative, so the square root has nothing to give. Under x = −3 the inside is exactly 0, and 0 = 0. From that column on, the inside is 0 or more and the root works: under x = −1 the inside is 4 and 4 = 2, and under x = 5 the inside is 16 and 16 = 4. The domain starts at the column where the inside reaches 0.

input xoutput 2x + 6−5−4−4−2−30−22−14516↓ evaluate: input given, read the output below it
The inside 2x + 6 turns from negative to 0 in the column under x = −3, and that column is where the domain…
Why the cube root is different

The square root answers one question: what number, 0 or more, squares to this? Squares are never negative, so the question has no answer for a negative inside. The cube root answers: what number cubes to this? Cubes come in every size and sign: 23 = 8, 03 = 0, (−2)3 = −8. So that question always has an answer. Earlier in this unit the square root appeared when you undid a quadratic on a cut domain, and the cube root appeared when you undid a cubic with no cut. The same difference shows up here: one domain has an edge, the other has none.

By moving it yourself

The figure starts on y = x−4. The curve begins at x = 4, where the inside is 0, and nothing is drawn to the left of 4. Slide h to another whole number and the starting point moves there. The domain always begins at the starting point: for y = x−h it is x ≥ h.

−6−4−2246−6−4−2246
Slide h and watch the starting point of the square root curve, which is the left edge of its domain. Move with it.
FunctionKind of rootCondition on the insideDomainInterval
xsquare (even)x≥0x≥0[0, ∞)
x−4square (even)x−4≥0x≥4[4, ∞)
2x+6square (even)2x+6≥0x≥−3[−3,∞)
10−2xsquare (even)10−2x≥0x≤5(−∞,5]
x3cube (odd)noneevery real number(−∞,∞)
∛(4x−7)cube (odd)noneevery real number(−∞,∞)

Part 1: The square root function

A square root function has x under a square root sign. The plainest one is f(x) = x. It asks: what number, 0 or more, squares to x? It can answer for 0 (the answer is 0) and for any positive number, but not for a negative one. So its graph has a starting point and runs off in one direction only. For ax+b with a ≠ 0, the starting point is the x that makes the inside equal 0.

  • Formula: f(x) = x, and more generally f(x) = ax+b
  • Domain of x: x ≥ 0, written [0, ∞)
  • Range of x: outputs of 0 or more, written [0, ∞)
  • Domain of ax+b: solve ax + b ≥ 0
  • Graph: starts at one point and curves away to one side, rising more slowly as it goes
  • Points on y = x: (0, 0), (1, 1), (4, 2), (9, 3)
(0, 0)(1, 1)(4, 2)(9, 3)domainrange−2246810−11234
The graph of y = x starts at (0, 0) and exists only for x ≥ 0; nothing is drawn to the left of 0.

Worked example

Domain first, then one output

For f(x) = x+9, find the domain. Then find f(7).

Show the steps and the answer
−9failsworks[−9, ∞)
The domain of x+9 begins at the filled dot at −9; the input 7 lies in the shaded part and −10 does not.
  1. Say what the question asks. First: find every x that f accepts. Second: the input is 7, find the output.The domain is the set of allowed inputs, and f(7) means put 7 where x is.
  2. Name the root: a square root with inside x+9. Write x+9≥0.The inside of a square root must be 0 or more.
  3. Subtract 9 from both sides: x≥−9.Subtracting the same number from both sides keeps an inequality true. This finds the edge: −9 is the one input that makes the inside exactly 0.
  4. Plug the edge back in: −9+9=0, and 0=0.The edge gives a real output, so it is included: [−9, ∞).
  5. Check that 7 is allowed: 7≥−9 is true.You may only evaluate a function at inputs in its domain.
  6. Find f(7): 7+9=16=4.7 + 9 = 16, and 4 is the number, 0 or more, whose square is 16.
Answer
  • Domain: x ≥ −9, or [−9, ∞)
  • f(7) = 4
Double-check Square the output: 42 = 16, and 16 is the inside 7 + 9. For the domain, test a number outside it: x = −10 gives −10 + 9 = −1, and −1 is not a real number, so −10 is correctly left out.
Tips and tricks
  • For ax+b with a ≠ 0, the starting point of the graph and the endpoint of the domain are the same x: the one that makes the inside 0.
  • The outputs of a square root are never negative. 16 is 4, not −4 and not both.

Part 2: The cube root function

A cube root function has x under a cube root sign. The plainest one is g(x) = x3. It asks: what number, cubed, gives x? That question always has exactly one answer. For 8 the answer is 2. For −8 the answer is −2, because (−2) · (−2) · (−2) = −8. For 0 the answer is 0. Since no input is ever turned away, the graph has no starting point. It is one unbroken curve that runs left and right forever.

  • Formula: g(x) = x3, and more generally g(x) = ∛(ax + b) with a ≠ 0
  • Domain: every real number, written (−∞, ∞)
  • Range: every real number, written (−∞, ∞)
  • Graph of x3: one unbroken S-shaped curve through (0, 0)
  • Points on y = x3: (−8, −2), (−1, −1), (0, 0), (1, 1), (8, 2)
(−8, −2)(−1, −1)(0, 0)(1, 1)(8, 2)domain−8−448−22
The graph of y = x3 continues on both sides of 0, so every x, negative ones included, has an output.

Worked example

A cube root takes a negative inside

For g(x) = ∛(x − 2), find the domain. Then find g(10) and g(−6).

Show the steps and the answer
(−6, −2)(2, 0)(10, 2)domain−8−44812−22
The graph of y = ∛(x − 2) passes through (−6, −2), (2, 0) and (10, 2) with no starting point and no gap.
  1. Say what the question asks. First: find every x that g accepts. Then: the input is 10, find the output, and the input is −6, find the output.The domain is the set of allowed inputs, and g(10) means put 10 where x is.
  2. Name the root: it is a cube root.A cube root is an odd root, and an odd root accepts negative, zero and positive insides.
  3. Write the domain: every real number, (−∞,∞). No inequality is needed.Whatever x you choose, x − 2 is some real number, and every real number has a cube root.
  4. Find g(10): (10−2)3=83=2.2 · 2 · 2 = 8.
  5. Find g(−6): (−6−2)3=(−8)3=−2.(−2) · (−2) · (−2) = −8. The first two negatives multiply to 4, and 4 · (−2) = −8.
Answer
  • Domain: every real number, or (−∞, ∞)
  • g(10) = 2
  • g(−6)=−2
Double-check Cube each output and compare with the inside. 23 = 8, and 10 − 2 = 8. (−2)3 = −8, and −6 − 2 = −8. Both match, and the second one shows a negative inside causing no trouble.
Tips and tricks
  • A cube root of a negative number is negative. A cube root of a positive number is positive. The sign passes straight through.
  • If you catch yourself writing inside ≥ 0 for a cube root, stop and test a negative inside such as ∛(−8) = −2.

5 How

  1. Find the root sign and name it: square root (even) or cube root (odd). These steps are for an inside of the form ax + b with a ≠ 0.
  2. If it is a cube root, stop: a cube root accepts every real number, so the domain is every real number, (−∞, ∞).
  3. If it is a square root, copy everything under the root sign and write: inside ≥ 0.
  4. Solve the inequality for x the way you solve an equation: undo the adding or subtracting first, then undo the multiplying. If you multiply or divide both sides by a negative number, flip the sign.
  5. Write the domain as an inequality and as an interval. When the root is not the bottom of a fraction, the endpoint gets a square bracket, because an inside of 0 is allowed: 0 = 0.
  6. Check with three numbers: the endpoint (the inside should be 0), one number in the domain (the inside should be positive), and one number outside it (the inside should be negative).

Finding the domain of a radical function

1
Is x under a root sign?
YesGo on to the next question.
NoThe root does not restrict x. A root of a plain number, such as 5, is a fixed number.
↓
2
Is it a cube root, or any other odd root?
YesStop. An odd root accepts every real number, so when nothing else limits x, as in ∛(ax + b), the domain is (−∞, ∞).
NoIt is a square root or another even root. Go on.
↓
3
Is the root the whole bottom of a fraction?
YesWrite inside > 0, because an inside of 0 would mean dividing by zero.
NoWrite inside ≥ 0.
↓
4
While solving, did you multiply or divide both sides by a negative number?
YesFlip the inequality sign at that step.
NoKeep the sign as it is.
  1. Name the root: square root (even) or cube root (odd). The inside is ax + b with a ≠ 0.
  2. Cube root: write (−∞, ∞) and stop.
  3. Square root: copy the whole inside and write inside ≥ 0.
  4. Solve for x. Undo adding or subtracting first, then undo the multiplying. Flip the sign if you divide or multiply by a negative number.
  5. Write the domain two ways: as an inequality and as an interval with a square bracket at the endpoint (a round one when the root is the whole bottom of a fraction).
  6. Test three numbers: the endpoint, one number in the domain, and one number outside it.

6 Worked example

Worked example

The domain of a square root function

Find the domain of f(x) = 2x+6.

Show the steps and the answer
−3failsworks[−3, ∞)
The test input 5 sits in the shaded domain of 2x+6 and the test input −4 sits outside it.
  1. Say what the question asks. Find the domain means: find every x you can put into f and get a real number out.The domain is the set of inputs the function accepts.
  2. Name the root. It is a square root, and the inside is 2x+6.A square root is an even root, so its inside may not be negative. A cube root would have no condition at all.
  3. Write the condition: 2x+6≥0.This sentence says the inside is 0 or more, which is exactly what the square root needs.
  4. Subtract 6 from both sides: 2x≥−6.Subtracting the same number from both sides keeps an inequality true, and it leaves the x term alone on the left.
  5. Divide both sides by 2: x≥−3.2 is positive, so the sign does not flip. This step finds the edge of the domain: −3 is the one input that makes the inside exactly 0, and every x to its right makes the inside positive.
  6. Plug the edge back in: 2·(−3)+6=−6+6=0, and 0=0.The edge gives a real output, so −3 belongs to the domain and gets a square bracket.
  7. Write the domain as an interval: [−3,∞).The square bracket at −3 says −3 is included. Infinity always gets a round bracket because it is never reached.
Answer
  • Domain: x ≥ −3
  • Interval notation: [−3, ∞)
Double-check Test a number on each side of −3. In the domain, x = 5: 2 · 5 + 6 = 16 and 16 = 4, a real number. Outside it, x = −4: 2 · (−4) + 6 = −2, and −2 is not a real number. The graph agrees: the curve starts at (−3, 0) and nothing is drawn to its left.

Ladder: from easy to exam-hard

Try each rung before you open its steps.

Rung 1 of 6

X minus a number

Find the domain of f(x) = x−4.

Show the steps and the answer
4[4, ∞)
The domain of x−4 starts at the filled dot at 4 and runs right forever.
  1. Say what the question asks: find every x that can go into f and give a real number out.The domain is the set of allowed inputs.
  2. Name the root: it is a square root with inside x−4.A square root is an even root, so its inside may not be negative.
  3. Write the condition: x−4≥0.The inside must be 0 or more.
  4. Add 4 to both sides: x≥4.Adding the same number to both sides keeps an inequality true. This finds the edge of the domain: 4 is the one input that makes the inside exactly 0.
  5. Plug the edge back in: 4−4=0, and 0=0.The edge works, so it gets a square bracket.
Answer
  • Domain: x ≥ 4
  • Interval notation: [4, ∞)
Double-check Test x = 13: 13 − 4 = 9 and 9 = 3, a real number. Test x = 3: 3 − 4 = −1, and −1 is not a real number. Inputs from 4 up work and inputs below 4 fail, which matches [4, ∞).

Rung 2 of 6

A number multiplying x

Find the domain of f(x) = 3x+12.

Show the steps and the answer
  1. Say what the question asks: find every x that f accepts.The domain is the set of allowed inputs.
  2. Name the root: a square root with inside 3x+12. Write 3x+12≥0.The inside of a square root must be 0 or more.
  3. Subtract 12 from both sides: 3x≥−12.Undo the adding first, so the x term stands alone.
  4. Divide both sides by 3: x≥−4.3 is positive, so the sign stays as it is. This finds the edge: −4 is the one input that makes the inside exactly 0.
  5. Plug the edge back in: 3·(−4)+12=−12+12=0, and 0=0.The edge gives a real output, so it is included.
Answer
  • Domain: x ≥ −4
  • Interval notation: [−4, ∞)
Double-check Test x = 8: 3 · 8 + 12 = 36 and 36 = 6, a real number. Test x = −5: 3 · (−5) + 12 = −3, and −3 is not a real number. That matches [−4, ∞).

Rung 3 of 6

A negative number multiplying x

Find the domain of f(x) = 10−2x.

Show the steps and the answer
5(−∞, 5]
The domain of 10−2x ends at the filled dot at 5 and runs left forever.
00252 < 5: 5 is further right−2−5times −1: −2 > −5, the order reversed
2 is less than 5, but −2 is greater than −5.
  1. Say what the question asks: find every x that f accepts.The domain is the set of allowed inputs.
  2. Name the root: a square root with inside 10−2x. Write 10−2x≥0.The inside of a square root must be 0 or more.
  3. Subtract 10 from both sides: −2x≥−10.Undo the adding first. The x term keeps its negative sign.
  4. Divide both sides by −2 and flip the sign: x≤5.Dividing by a negative number reverses the order of numbers, so ≥ becomes ≤. And −10 ÷ (−2) = 5, since a negative divided by a negative is positive. This finds the edge: 5 is the one input that makes the inside exactly 0.
  5. Plug the edge back in: 10−2·5=0, and 0=0.The edge gives a real output, so it is included.
  6. Write the interval: (−∞,5].x ≤ 5 means everything to the left of 5, with 5 included. The domain runs left this time.
Answer
  • Domain: x ≤ 5
  • Interval notation: (−∞, 5]
Double-check Solve a second way that needs no flip: add 2x to both sides of 10 − 2x ≥ 0 to get 10 ≥ 2x, then divide by 2 to get 5 ≥ x, which is the same as x ≤ 5. Test x = 3: 10 − 6 = 4 and 4 = 2. Test x = 6: 10 − 12 = −2, and −2 is not a real number.

Rung 4 of 6

A cube root

Find the domain of g(x) = ∛(4x − 7).

Show the steps and the answer
  1. Say what the question asks: find every x that g accepts.The domain is the set of allowed inputs.
  2. Name the root: it is a cube root.The kind of root decides whether there is any condition at all.
  3. Write no inequality. The domain is every real number.A cube root accepts negative, zero and positive insides, and 4x − 7 is a real number for every x.
  4. Show it on a negative inside. At x=−5: 4·(−5)−7=−27, and (−27)3=−3.(−3) · (−3) · (−3) = −27, so even this negative inside has a cube root.
Answer
  • Domain: every real number
  • Interval notation: (−∞, ∞)
Double-check Try insides of all three kinds. Positive: x = 2 gives 4 · 2 − 7 = 1 and 13 = 1. Zero: x = 74 gives 7 − 7 = 0 and 03 = 0. Negative: x = 0 gives −7 and ∛(−7) ≈ −1.91 (rounded), a real number. No input fails.

Rung 5 of 6

An endpoint that is a fraction

Find the domain of f(x) = 5−4x.

Show the steps and the answer
1.25(−∞, 1.25]
The domain of 5−4x ends at the filled dot at 1.25, which is 54, and runs left.
  1. Say what the question asks: find every x that f accepts.The domain is the set of allowed inputs.
  2. Name the root: a square root with inside 5−4x. Write 5−4x≥0.The inside of a square root must be 0 or more.
  3. Subtract 5 from both sides: −4x≥−5.Undo the adding first.
  4. Divide both sides by −4 and flip the sign: x≤54.Dividing by a negative flips ≥ to ≤. And −5 ÷ (−4) = 54, positive because a negative divided by a negative is positive. This finds the edge of the domain.
  5. Plug the edge back in: 4·54=5, so 5−5=0, and 0=0.Four copies of five fourths make 5. The edge gives a real output, so it is included.
  6. Write the interval. 54 is exactly 1.25, so the domain is (−∞,1.25].An endpoint does not have to be a whole number. It is whatever x makes the inside 0.
Answer
  • Domain: x ≤ 54 (exactly 1.25)
  • Interval notation: (−∞, 1.25]
Double-check Test x = 1, which is less than 1.25: 5 − 4 = 1 and 1 = 1. Test x = −1: 5 + 4 = 9 and 9 = 3. Test x = 2, which is more than 1.25: 5 − 8 = −3, and −3 is not a real number. That matches x ≤ 54.

Rung 6 of 6

Numbers outside the root

Find the domain of h(x) = 32x−9 + 1.

Show the steps and the answer
  1. Say what the question asks: find every x that h accepts.The domain is the set of allowed inputs.
  2. Find the root and mark its inside: only 2x−9 is under the root sign. The 3 in front and the +1 at the end are outside.Only what sits under the root sign has to be 0 or more.
  3. Ignore the 3 and the +1 for the domain.Multiplying by 3 and adding 1 work on every real number, so they turn no input away.
  4. Write the condition: 2x−9≥0.The inside of a square root must be 0 or more.
  5. Add 9 to both sides: 2x≥9. Then divide both sides by 2: x≥92.2 is positive, so there is no flip. This finds the edge: 92, which is exactly 4.5, is the one input that makes the inside 0.
  6. Plug the edge back in: 2·4.5−9=0, so h(4.5)=3·0+1=3·0+1=1.The edge gives a real output, so it is included.
Answer
  • Domain: x ≥ 92 (exactly 4.5)
  • Interval notation: [4.5, ∞)
Double-check Test x = 9: 2 · 9 − 9 = 9, 9 = 3, and h(9) = 3 · 3 + 1 = 10, a real number. Test x = 4: 2 · 4 − 9 = −1, and −1 is not a real number, so 4 is correctly left out.

Counterexamples

What it is not, and the tempting wrong moves.

✗ Not this: For f(x) = x−5, the domain is x ≥ 0.
The condition is on the inside, not on x. The input 2 is 0 or more, yet 2 − 5 = −3 and −3 is not a real number.
✓ Instead: Write x − 5 ≥ 0, so x ≥ 5. The domain is [5, ∞). Test x = 9: 4 = 2.
✗ Not this: For f(x) = 12−3x: 12 − 3x ≥ 0, so −3x ≥ −12, so x ≥ 4.
The last step divided by −3 without flipping the sign. Test x = 6, which the wrong answer allows: 12 − 18 = −6, and −6 is not a real number.
✓ Instead: Dividing by −3 flips the sign: x ≤ 4. The domain is (−∞, 4]. Test x = 1: 12 − 3 = 9 and 9 = 3.
✗ Not this: For g(x) = ∛(x − 2), the inside must be 0 or more, so the domain is x ≥ 2.
The rule inside ≥ 0 belongs to square roots and other even roots. It does not apply to a cube root. The input −6 gives ∛(−8) = −2, a real number.
✓ Instead: A cube root takes every real number. The domain is (−∞, ∞).
✗ Not this: For k(x) = 1x−5, the domain is x ≥ 5, the same as for x−5.
This is a case where the usual ≥ does not apply. At x = 5 the root is 0 = 0, and then the fraction asks for 1 ÷ 0, which has no answer.
✓ Instead: When the root is the whole bottom of a fraction, write inside > 0. Here x − 5 > 0, so x > 5 and the domain is (5, ∞). Test x = 9: 14 = 12.

Tips and tricks

  • Memory cue: even roots are picky, odd roots are open. A square root is an even root. A cube root is an odd root.
  • Copy the whole inside before you write ≥ 0. The condition is never about x alone unless the inside is x alone.
  • For a square root (or any even root) of ax + b with a ≠ 0, the endpoint of the domain is the x that makes the inside 0, and it is where the graph starts.
  • For a square root, if the x term in the inside is negative, as in 10 − 2x, expect the domain to run left: x ≤ a number.
  • Check every square-root answer with three numbers: the endpoint, one number in the domain, and one outside it. A cube root has no endpoint and no number outside its domain: every x works. It takes less time than redoing the algebra.
  • On paper, stretch the root bar over the whole inside. 7−x and 7 − x are different: the second is the root of 7, then minus x.
  • You have met both roots already in this unit: the square root as the inverse function of a quadratic on a restricted domain, and the cube root as the inverse function of a cubic.

7 The trap

Writing x ≥ 0 out of habit. The condition belongs to the whole inside, not to x alone. For f(x) = 2x+6, the answer x ≥ 0 throws away every input from −3 up to, but not including, 0, and those inputs work: x = −1 gives 4 = 2. To avoid it, copy everything under the root sign first, write ≥ 0 after it, and only then solve.

Flash cards

Say the answer out loud, then flip.

Why does the square root guard turn away every negative inside?
  • Squaring never gives a negative: 32 = 9, (−3)2 = 9 and 02 = 0.
  • So no real number squares to a negative, and the inside must be 0 or more.
For f(x) = 2x+6, what do you write first, and what domain does it give?
  • Copy the whole inside and write 2x + 6 ≥ 0.
  • Solve: 2x ≥ −6, so x ≥ −3.
  • Domain: [−3, ∞).
  • Not x ≥ 0: x = −1 gets past the guard, because 4 = 2.
What is the domain of a cube root function, and why is its gate open?
  • Every real number, (−∞, ∞).
  • Cubing keeps the sign: 23 = 8 and (−2)3 = −8, so ∛(−8) = −2 is a real answer.
In the domain of a square root function, why does the endpoint get a square bracket?
  • Because an inside of exactly 0 is allowed: 0 = 0.
  • The guard turns away only negatives, so the endpoint is in the domain.

Lesson 5 of 6

Inverses of rational functions

Subject: MathematicsLeans on: The name f⁻¹Leans on: Zero in the bottomLeans on: Clearing a fractionLeans on: DistributingLeans on: Moving a termLeans on: Factoring out a common letterLeans on: Writing every number except oneOpen this lesson
Leans on 0Comes back in 0

1 In plain words

A rational function is a fraction whose top and bottom are polynomials. This lesson works with the kind ax+bcx+d that has x in the bottom and does not cancel down to a constant, such as 2x+1x−3. Picture a machine that uses your number twice, once upstairs and once downstairs, then divides. To run it backward you swap x and y and solve for y, as you did before. The new problem: after the swap, y sits in two places. It is like one pair of socks left in two drawers. Bring them into one drawer first. Here that means: move every term with y to one side, pull y out as a common factor, then divide. The machine has one number it cannot take in and one number it never gives out. The inverse function trades those two numbers.

2 Picture it

(4, 9)(10, 3)−44812−8−44812(4, 9)(10, 3)−44812−8−44812
The graph of f(x) = 2x+1x−3 never touches the dashed line x = 3 (the input it cannot take) or the dashed…

Reminder

Reminder: The name f⁻¹, Zero in the bottom, Clearing a fraction, Distributing, Moving a term, Factoring out a common letter, Writing every number except one
  • The name f⁻¹. f⁻¹ is the name of the inverse function. The small −1 is not a power: f⁻¹(x) does not mean 1f(x).
  • Zero in the bottom. A fraction with 0 in the bottom has no value, so any input that makes the bottom 0 is left out of the domain: for 5x−4, leave out x = 4.
  • Clearing a fraction. Multiplying a fraction by its own bottom leaves only the top: 62 · 2 = 6, and 6y−1 · (y − 1) = 6.
  • Distributing. The number or letter outside multiplies every term inside: x(y − 3) = xy − 3x.
  • Moving a term. Do the same thing to both sides. To move 2y off the right of xy = 2y + 1, subtract 2y from both sides: xy − 2y = 1.
  • Factoring out a common letter. If every term on a side contains y, write y once outside brackets: xy − 2y = y(x − 2). Distribute back to check.
  • Writing every number except one. x ≠ 3 in interval notation is (−∞, 3) ∪ (3, ∞): everything left of 3, joined with everything right of 3.

3 Why

The inverse function answers one question: which input produced this output. Swapping x and y writes that question down, and solving for y answers it. Factoring is needed because xy − 2y holds y twice, and no single division can free a letter that appears twice. Written as y(x − 2), y appears once, times a bracket, and division undoes that. Domain and range trade because every pair (input, output) of f, read backward, is a pair of the inverse. So the inputs f accepts are the outputs its inverse returns, and the other way around.

4 The rule

For f(x)=ax+bcx+d with c≠0 and ad−bc≠0, swap and solve gives f−1(x)=−dx+bcx−a.

Domain of f: x≠−dc. Range of f: y≠ac. Domain of f−1= range of f. Range of f−1= domain of f.

The same idea, five ways

Say it

f inverse of x equals three x plus one, over x minus two.

Write it

The inverse function of f(x) = 2x+1x−3 is f⁻¹(x) = 3x+1x−2, and the number each one cannot take in is the number the other never gives out.

In math
  • f(x)=2x+1x−3
  • f⁻¹(x)=3x+1x−2
  • Domain of f = range of f⁻¹: x ≠ 3, (−∞, 3) ∪ (3, ∞), {x | x ≠ 3}
  • Range of f = domain of f⁻¹: y ≠ 2, (−∞, 2) ∪ (2, ∞), {y | y ≠ 2}
Like

A one-way street with one closed entrance and one closed exit. Drive it in reverse and the closed entrance becomes the closed exit.

See it

f(x)=2x+1x−3

f⁻¹(x)=3x+1x−2

f: in ≠ 3, out ≠ 2

f⁻¹: in ≠ 2, out ≠ 3

The same idea, other ways

As a picture

The graph of f(x) = 2x+1x−3 has two dashed guide lines it never touches: an up-and-down line at x = 3, the input f cannot take, and a level line at y = 2, the output f never gives. The graph shown here is the inverse function f⁻¹(x) = 3x+1x−2. It has the same two numbers with their jobs traded: the up-and-down line is at x = 2 and the level line is at y = 3. The points trade too: (4, 9) on f becomes (9, 4) here.

(9, 4)(3, 10)−8−44812−44812
The inverse function never touches x = 2 or y = 3, the two left-out numbers of f with their jobs traded.
With small numbers

The table lists four inputs of f(x) = 2x+1x−3 and their outputs. Read a column from top to bottom and you are running f: 4 goes in, 9 comes out. Read the same column from bottom to top and you are running the inverse function: 9 goes in, 4 comes out. The formula agrees: f⁻¹(9) = 3(9)+19−2 = 287 = 4. However long you make this table, 3 can never appear in the input row and 2 can never appear in the output row. The 3 you see in the output row is allowed: 3 is a fine output, only a forbidden input.

input xoutput f(x)−412−549103↓ evaluate: input given, read the output below it
Top to bottom is f, bottom to top is the inverse function: the column under 4 says f(4) = 9 and f⁻¹(9) = 4.
Why it must be true

Ask f to produce the output 2: 2x+1x−3 = 2. Multiply both sides by (x − 3): 2x + 1 = 2x − 6. Subtract 2x from both sides: 1 = −6. That is false, so no input gives 2. The inverse function takes outputs of f as its inputs, so it has nothing to hand back for 2. Its formula agrees: the bottom x − 2 is zero at x = 2.

As a rule of thumb

Swap, clear, gather, factor, divide. Swap x and y. Clear the fraction by multiplying by the bottom. Gather the y terms on one side. Factor y out. Divide by the bracket. Then read each function's forbidden input from its own bottom, and hand that number to the other function as its missing output.

Swap: x = 2y+1y−3

Clear: x(y − 3) = 2y + 1

Gather: xy − 2y = 3x + 1

Factor: y(x − 2) = 3x + 1

Divide: y = 3x+1x−2

FunctionInput it cannot takeOutput it never gives
f(x)=2x+1x−332
f⁻¹(x)=3x+1x−223

5 How

  1. Find the input f cannot take: set the bottom of f equal to 0 and solve. Leave that number out of the domain of f.
  2. Write y in place of f(x).
  3. Swap x and y everywhere.
  4. Multiply both sides by the whole bottom, written in brackets, so the fraction disappears.
  5. Distribute to remove the brackets.
  6. Gather: move every term that contains y to one side and every term without y to the other side.
  7. Factor y out of the side that holds it, so y appears once, times a bracket.
  8. Divide both sides by that bracket. Rename y as f⁻¹(x).
  9. Find the input f⁻¹ cannot take: set the bottom of f⁻¹ equal to 0 and solve. That number is left out of the domain of f⁻¹ and out of the range of f.
  10. Trade: the range of f⁻¹ is the domain of f from step 1.
  11. Check with one number: pick an input, run f, feed the output to f⁻¹, and confirm you get the input back.

Inverting a rational function and matching domain and range

1
After the swap, is y in the bottom of a fraction?
YesMultiply both sides by the whole bottom, in brackets, then distribute.
NoGo on.
↓
2
Is y in more than one term?
YesMove every y term to one side and everything else to the other, then factor y out.
NoMove the terms without y to the other side.
↓
3
Is y multiplied by a bracket or a number?
YesDivide both sides by it. The result is f⁻¹(x).
Noy is alone. The solving is done.
↓
4
Does your f⁻¹ have x in a bottom?
YesSet that bottom = 0 and solve. Leave that x out of the domain of f⁻¹. The same number is left out of the range of f.
NoRecheck the algebra. When f has x in its bottom, its inverse has x in a bottom too.
  1. Set the bottom of f equal to 0 and solve. Leave that input out of the domain of f.
  2. Write y for f(x), then swap x and y.
  3. Clear the fraction: multiply both sides by the bottom, in brackets, and distribute.
  4. Gather every y term on one side and everything else on the other.
  5. Factor y out, then divide by the bracket. Name the result f⁻¹(x).
  6. Set the bottom of f⁻¹ equal to 0 and solve. Leave that input out of the domain of f⁻¹.
  7. Trade: range of f = domain of f⁻¹, and range of f⁻¹ = domain of f.
  8. Check one number through f and back through f⁻¹.

6 Worked example

Worked example

Inverse of a line over a line, with domain and range

Find the inverse function of f(x) = 2x+1x−3. Then give the domain and range of f and of f⁻¹.

Show the steps and the answer
=x(y − 3)2y + 1do the same thing to both sides
After multiplying by the bottom, the fraction is gone and y sits on both pans.
3(−∞, 3) ∪ (3, ∞)
The domain of f is every number except 3, shown as an open dot at 3.
2(−∞, 2) ∪ (2, ∞)
The domain of f⁻¹ is every number except 2, which is also the range of f.
  1. Say what the question asks. f takes an input x and gives an output. The inverse function takes that output and hands back the input. Find its formula, then list which numbers each function takes in (domain) and gives out (range).Knowing the goal in plain words tells you when you are finished: one formula and four sets of numbers.
  2. Find the input f cannot take. Set the bottom equal to 0: x−3=0, so x=3. Domain of f: x≠3.This step finds the one input that would put 0 in the bottom, so we can throw it out. Plug it in to see: 2(3)+13−3 = 70, which has no value.
  3. Write y for f(x): y=2x+1x−3.y is a shorter name for the output, and it makes the swap possible to write.
  4. Swap x and y: x=2y+1y−3.The inverse function trades the jobs of input and output. After the swap, x is the old output and y is the old input we want back.
  5. Multiply both sides by (y − 3): x(y−3)=2y+1.y is trapped in a bottom. Multiplying a fraction by its own bottom leaves only the top, so the fraction disappears. The brackets make x multiply the whole bottom.
  6. Distribute on the left: xy−3x=2y+1.x multiplies both things inside the brackets: x · y and x · (−3).
  7. Gather the y terms. Subtract 2y from both sides, then add 3x to both sides: xy−2y=3x+1.y sits in two terms, xy and 2y. Put both on the left and everything without y on the right, so y can be pulled out.
  8. Factor y out of the left side: y(x−2)=3x+1.Both left terms contain y: xy = y · x and 2y = y · 2. Check by distributing back: y(x − 2) = xy − 2y.
  9. Divide both sides by (x − 2): y=3x+1x−2.y is now multiplied by one bracket, and division undoes multiplication. The division is allowed only when x − 2 is not 0, which is where the next restriction comes from.
  10. Rename y: f−1(x)=3x+1x−2.The y we solved for is the old input written in terms of the old output. That is the inverse function.
  11. Find the input f−1 cannot take. Set its bottom equal to 0: x−2=0, so x=2. Domain of f−1: x≠2.This finds the one input that would put 0 in the bottom of the inverse. Plug it in to see: 3(2)+12−2 = 70, which has no value.
  12. Trade the sets. Range of f= domain of f−1: y≠2. Range of f−1= domain of f: y≠3.The outputs of f are exactly the inputs of f⁻¹, and the outputs of f⁻¹ are exactly the inputs of f.
Answer
  • f⁻¹(x)=3x+1x−2
  • Domain of f: x ≠ 3, or (−∞, 3) ∪ (3, ∞)
  • Range of f: y ≠ 2, or (−∞, 2) ∪ (2, ∞)
  • Domain of f⁻¹: x ≠ 2, or (−∞, 2) ∪ (2, ∞)
  • Range of f⁻¹: y ≠ 3, or (−∞, 3) ∪ (3, ∞)
Double-check Run a number through both machines. f(4) = 2(4)+14−3 = 91 = 9. Then f⁻¹(9) = 3(9)+19−2 = 287 = 4, the input we started with. A second pair: f(10) = 217 = 3 and f⁻¹(3) = 101 = 10. For the range, ask f for the output 2 directly: 2x+1x−3 = 2 means 2x + 1 = 2x − 6, so 1 = −6, which is false. No input gives 2, as claimed.

Ladder: from easy to exam-hard

Try each rung before you open its steps.

Rung 1 of 6

X only in the bottom

Find the inverse function of f(x) = 6x−1, with the domain and range of both functions.

Show the steps and the answer
(4, 2)−448−8−448
The graph of f(x) = 6x−1 never touches x = 1 or y = 0, the input and output it leaves out.
  1. Say what the question asks: a formula that turns each output of f back into its input, plus the numbers each function leaves out.The goal in plain words tells you what a finished answer contains.
  2. Set the bottom equal to 0: x−1=0, so x=1. Domain of f: x≠1.This finds the one input that would put 0 in the bottom. Plug it in: 61−1 = 60, which has no value.
  3. Write y=6x−1, then swap: x=6y−1.Swapping makes x the old output and y the old input we want to recover.
  4. Multiply both sides by (y − 1): x(y−1)=6.Multiplying by the bottom removes the fraction and frees y.
  5. Distribute: xy−x=6.x multiplies both y and −1.
  6. Add x to both sides: xy=x+6.y is in only one term, so there is nothing to gather. Move the term without y away from it.
  7. Divide both sides by x: y=x+6x. So f−1(x)=x+6x.y was multiplied by x, and division undoes that.
  8. Set the bottom of f−1 equal to 0: x=0. Domain of f−1: x≠0. Trade: range of f is y≠0, and range of f−1 is y≠1.This finds the one input the inverse cannot take: 0+60 = 60 has no value. It makes sense that f never gives 0, since a fraction with 6 on top is never 0.
Answer
  • f⁻¹(x)=x+6x
  • f: domain x ≠ 1, range y ≠ 0
  • f⁻¹: domain x ≠ 0, range y ≠ 1
Double-check f(4) = 64−1 = 63 = 2. Then f⁻¹(2) = 2+62 = 82 = 4, the input we started with.

Rung 2 of 6

X on top and bottom, so y must be gathered

Find the inverse function of f(x) = x+5x−2, with the domain and range of both functions.

Show the steps and the answer
  1. Say what the question asks: the function that undoes f, and the numbers each one leaves out.Same goal as before. The new feature is x in two places.
  2. Set the bottom equal to 0: x−2=0, so x=2. Domain of f: x≠2.This finds the input that would put 0 in the bottom: 2+52−2 = 70, which has no value.
  3. Write y=x+5x−2, then swap: x=y+5y−2.The swap trades input and output.
  4. Multiply both sides by (y − 2): x(y−2)=y+5.This clears the fraction.
  5. Distribute: xy−2x=y+5.x multiplies both y and −2.
  6. Subtract y from both sides, then add 2x to both sides: xy−y=2x+5.y is in two terms, xy and y. Gather them on the left and send everything else to the right.
  7. Factor y out: y(x−1)=2x+5.xy = y · x and y = y · 1, so a 1 stays in the bracket. Distribute back to check: y · x − y · 1 = xy − y.
  8. Divide by (x−1): y=2x+5x−1. So f−1(x)=2x+5x−1.Division undoes the multiplication by the bracket.
  9. Set the bottom of f−1 equal to 0: x−1=0, so x=1. Domain of f−1: x≠1. Trade: range of f is y≠1, and range of f−1 is y≠2.This finds the input the inverse cannot take: 2(1)+51−1 = 70 has no value. Each function's left-out input is the other's left-out output.
Answer
  • f⁻¹(x)=2x+5x−1
  • f: domain x ≠ 2, range y ≠ 1
  • f⁻¹: domain x ≠ 1, range y ≠ 2
Double-check f(3) = 3+53−2 = 81 = 8. Then f⁻¹(8) = 2(8)+58−1 = 217 = 3, the input we started with.

Rung 3 of 6

A number in front of x in the bottom

Find the inverse function of f(x) = 3x−22x+1, with the domain and range of both functions.

Show the steps and the answer
−0.5(−∞, −0.5) ∪ (−0.5, ∞)
The domain of f is every number except −12, the one input that makes 2x + 1 equal 0.
  1. Say what the question asks: the function that undoes f, and the numbers each one leaves out. Expect fractions among those numbers.The 2 in front of x in the bottom means the forbidden input will not be a whole number.
  2. Set the bottom equal to 0: 2x+1=0, so 2x=−1 and x=−12. Domain of f: x≠−12.This finds the input that makes the bottom 0. Plug it in: 2(−12) + 1 = −1 + 1 = 0.
  3. Write y=3x−22x+1, then swap: x=3y−22y+1.The swap trades input and output.
  4. Multiply both sides by (2y + 1): x(2y+1)=3y−2.This clears the fraction.
  5. Distribute: 2xy+x=3y−2.x multiplies both 2y and 1.
  6. Subtract 3y from both sides, then subtract x from both sides: 2xy−3y=−x−2.The y terms are 2xy and 3y. Gather them on the left. The x that has no y moves to the right.
  7. Factor y out: y(2x−3)=−x−2.2xy = y · 2x and 3y = y · 3. Distribute back to check: y · 2x − y · 3 = 2xy − 3y.
  8. Divide by (2x−3): y=−x−22x−3. So f−1(x)=−x−22x−3.Division undoes the multiplication by the bracket.
  9. Set the bottom of f−1 equal to 0: 2x−3=0, so 2x=3 and x=32. Domain of f−1: x≠32. Trade: range of f is y≠32, and range of f−1 is y≠−12.This finds the input the inverse cannot take. Plug it in: 2(32) − 3 = 3 − 3 = 0 in the bottom.
Answer
  • f⁻¹(x) = −x−22x−3, which can also be written x+23−2x by multiplying top and bottom by −1
  • f: domain x ≠ −12, range y ≠ 32
  • f⁻¹: domain x ≠ 32, range y ≠ −12
Double-check f(−1) = 3(−1)−22(−1)+1 = −5−1 = 5. Then f⁻¹(5) = −5−22(5)−3 = −77 = −1, the input we started with. The shortcut agrees on the range: the numbers in front of x are 3 on top and 2 on the bottom, and 32 is the missing output.

Rung 4 of 6

A negative x term on top

Find the inverse function of f(x) = 5−2xx+4, with the domain and range of both functions.

Show the steps and the answer
  1. Say what the question asks: the function that undoes f, and the numbers each one leaves out. Watch the signs when gathering.The −2x on top turns into −2y after the swap, and it will change sign when it crosses the equals sign.
  2. Set the bottom equal to 0: x+4=0, so x=−4. Domain of f: x≠−4.This finds the input that would put 0 in the bottom: 5−2(−4)−4+4 = 130, which has no value.
  3. Write y=5−2xx+4, then swap: x=5−2yy+4.The swap trades input and output.
  4. Multiply both sides by (y + 4): x(y+4)=5−2y.This clears the fraction.
  5. Distribute: xy+4x=5−2y.x multiplies both y and 4.
  6. Add 2y to both sides, then subtract 4x from both sides: xy+2y=5−4x.The y terms are xy and −2y. Adding 2y to both sides brings it to the left as + 2y.
  7. Factor y out: y(x+2)=5−4x.xy = y · x and 2y = y · 2. Distribute back to check: y · x + y · 2 = xy + 2y.
  8. Divide by (x+2): y=5−4xx+2. So f−1(x)=5−4xx+2.Division undoes the multiplication by the bracket.
  9. Set the bottom of f−1 equal to 0: x+2=0, so x=−2. Domain of f−1: x≠−2. Trade: range of f is y≠−2, and range of f−1 is y≠−4.This finds the input the inverse cannot take: 5−4(−2)−2+2 = 130 has no value.
Answer
  • f⁻¹(x)=5−4xx+2
  • f: domain x ≠ −4, range y ≠ −2
  • f⁻¹: domain x ≠ −2, range y ≠ −4
Double-check f(−3) = 5−2(−3)−3+4 = 111 = 11. Then f⁻¹(11) = 5−4(11)11+2 = −3913 = −3, the input we started with.

Rung 5 of 6

Using the inverse to find a range

Find the range of g(x) = 4x−12x+6.

Show the steps and the answer
2(−∞, 2) ∪ (2, ∞)
The range of g is every number except 2, shown as an open dot at 2.
  1. Say what the question asks. The range is the set of all outputs of g. Outputs are hard to read from a formula, but the outputs of g are the inputs of its inverse function, and inputs can be read from a bottom. So find the inverse and take its domain.Range of g = domain of g⁻¹. A domain question is one you already know how to answer.
  2. Write y=4x−12x+6, then swap: x=4y−12y+6.The swap trades input and output.
  3. Multiply both sides by (2y + 6): x(2y+6)=4y−1.This clears the fraction.
  4. Distribute: 2xy+6x=4y−1.x multiplies both 2y and 6.
  5. Subtract 4y from both sides, then subtract 6x from both sides: 2xy−4y=−6x−1.The y terms are 2xy and 4y. Gather them on the left and send the rest to the right.
  6. Factor y out: y(2x−4)=−6x−1.2xy = y · 2x and 4y = y · 4. Distribute back to check: y · 2x − y · 4 = 2xy − 4y.
  7. Divide by (2x−4): y=−6x−12x−4. So g−1(x)=−6x−12x−4.Division undoes the multiplication by the bracket.
  8. Set the bottom of g−1 equal to 0: 2x−4=0, so 2x=4 and x=2. Domain of g−1: x≠2. So the range of g is y≠2.This finds the one input the inverse cannot take. Plug it in: 2(2) − 4 = 0 in the bottom. That input of g⁻¹ is the one output g never gives.
Answer
  • Range of g: every real number except 2
  • In interval notation: (−∞, 2) ∪ (2, ∞)
Double-check Ask g for the output 2 directly: 4x−12x+6 = 2 means 4x − 1 = 2(2x + 6) = 4x + 12, so −1 = 12, which is false. No input gives 2. Another output, such as 1, can be reached: 4x − 1 = 2x + 6 gives 2x = 7, so x = 72, and g(72) = 14−17+6 = 1313 = 1. The shortcut agrees: 42 = 2.

Rung 6 of 6

The gotcha, a function that is its own inverse

Find the inverse function of f(x) = x+5x−1, with the domain and range of both functions.

Show the steps and the answer
(2, 7)(7, 2)−448−448
Both (2, 7) and (7, 2) lie on the same curve.
  1. Say what the question asks: the function that undoes f, and the numbers each one leaves out.Same goal. The surprise comes at the end.
  2. Set the bottom equal to 0: x−1=0, so x=1. Domain of f: x≠1.This finds the input that would put 0 in the bottom: 1+51−1 = 60, which has no value.
  3. Write y=x+5x−1, then swap: x=y+5y−1.The swap trades input and output.
  4. Multiply both sides by (y−1) and distribute: xy−x=y+5.Clearing the fraction gives x(y − 1) = y + 5, and x multiplies both y and −1.
  5. Subtract y from both sides, then add x to both sides: xy−y=x+5.Gather the y terms, xy and y, on the left.
  6. Factor y out: y(x−1)=x+5.xy = y · x and y = y · 1. Distribute back to check: xy − y.
  7. Divide by (x−1): y=x+5x−1. So f−1(x)=x+5x−1.Division undoes the multiplication by the bracket.
  8. Notice that f−1 has the same formula as f. Nothing went wrong. Domain of f−1: x≠1. Range of f: y≠1. Range of f−1: y≠1.Some functions undo themselves. Here the forbidden input and the missing output are both 1, so trading them changes nothing. The corner check agrees: 1 and −1 trade places and change sign, which gives 1 and −1 again.
Answer
  • f⁻¹(x) = x+5x−1, the same formula as f
  • f: domain x ≠ 1, range y ≠ 1
  • f⁻¹: domain x ≠ 1, range y ≠ 1
Double-check f(2) = 2+52−1 = 71 = 7. Then f⁻¹(7) = 7+57−1 = 126 = 2, the input we started with. A second pair: f(3) = 82 = 4 and f⁻¹(4) = 93 = 3. Running the same formula twice returns the starting number.

Counterexamples

What it is not, and the tempting wrong moves.

✗ Not this: Flipping the fraction: f⁻¹(x) = x−32x+1.
Turning a fraction upside down gives 1 divided by f(x), which is a different thing from undoing f. Test it: f(4) = 9, but 9−32(9)+1 = 619, which is not 4.
✓ Instead: Swap x and y and solve: f⁻¹(x) = 3x+1x−2. Then f⁻¹(9) = 287 = 4.
✗ Not this: Saying the domain of f⁻¹ is x ≠ 3 because the domain of f is x ≠ 3.
Each function has its own bottom. The inverse accepts 3 without trouble: f⁻¹(3) = 3(3)+13−2 = 101 = 10. Its bottom is x − 2, which is zero at 2.
✓ Instead: Domain of f⁻¹: x ≠ 2. The number 3 moves to the range of f⁻¹: y ≠ 3.
✗ Not this: Factoring xy − y as y(x) or as y(x − 0), dropping the second term.
The lone y is y · 1. Pulling y out leaves a 1 behind. Distribute to test: y(x) = xy, which has lost the − y.
✓ Instead: xy − y = y(x − 1). Distribute back: y · x − y · 1 = xy − y.
✗ Not this: Looking for the inverse of f(x) = 2x+4x+2 by swap and solve.
The rule does not apply. The top is 2 times the bottom, since 2(x + 2) = 2x + 4, so the fraction equals 2 for every allowed x. Many inputs share one output, so this is not a one-to-one function. Here ad − bc = 2 · 2 − 4 · 1 = 0, the case the rule excludes.
✓ Instead: This function has no inverse function. Before inverting, glance at whether the top is a multiple of the bottom.

Tips and tricks

  • Five verbs carry the whole method: swap, clear, gather, factor, divide.
  • A ten-second check, not something to memorize: compare f(x) = ax+bcx+d with your answer. The two corner numbers a and d should have traded places and changed sign, while b and c stayed put. For 2x+1x−3: 2 and −3 trade and flip sign, giving 3x+1x−2.
  • Do not memorize the formula in the rule. Swap and solve rebuilds it every time, and the steps are the skill.
  • Shortcut for the output f never gives: divide the number in front of x on top by the number in front of x on the bottom. For 2x+1x−3 that is 21 = 2. If the top has no x, the missing output is 0.
  • For a check with whole numbers, pick the input that makes the bottom equal 1. For x − 3 that is x = 4.
  • No cut is needed here. Swap and solve gives exactly one y for each x, so each output comes from one input. Unlike a parabola, there is no sign to choose.
  • Always write the bottom in brackets when you multiply by it: x(y − 3), never xy − 3.

7 The trap

After the fraction is cleared, y sits in two terms on opposite sides, as in xy − 3x = 2y + 1. The common mistake is to divide right away and write y = 2y+3x+1x. That is not solved, because y is still on the right. Gather every y term on one side first, factor y out, and only then divide.

Flash cards

Say the answer out loud, then flip.

After you swap x and y in a rational function and clear the fraction, y sits in two terms. What three moves free it?
  • Gather every term with y on one side.
  • Factor y out, so y appears once, times a bracket.
  • Divide both sides by that bracket.
  • It is like one pair of socks left in two drawers: bring them into one drawer first.
From xy − 3x = 2y + 1, someone writes y = 2y+3x+1x. Why is that not a finished answer, and what is the fix?
  • y is still on the right side, so nothing is solved. No single division can free a letter that appears twice.
  • Gather: xy − 2y = 3x + 1.
  • Factor: y(x − 2) = 3x + 1.
  • Divide: y = 3x+1x−2.
For f(x) = 2x+1x−3, the inverse function is f⁻¹(x) = 3x+1x−2. Which number is left out of the domain of each, and how do those two numbers trade?
  • Domain of f: x ≠ 3, because 3 makes the bottom x − 3 equal 0.
  • Domain of f⁻¹: x ≠ 2, because 2 makes the bottom x − 2 equal 0.
  • The trade: 2 is the one number f never gives out, so the range of f is y ≠ 2. And 3 is the one number f⁻¹ never gives out, so the range of f⁻¹ is y ≠ 3.
How do you check an inverse function with one number? Try it on f(x) = 2x+1x−3 and f⁻¹(x) = 3x+1x−2 with the input 4.
  • Pick an input, run f, feed the output to f⁻¹, and confirm the input comes back.
  • f(4)=2·4+14−3=91=9.
  • f⁻¹(9)=3·9+19−2=287=4.
  • The 4 came back, so the inverse passes the check.

Lesson 6 of 6

A function and its inverse undo each other

Subject: MathematicsLeans on: Reading f(g(x))Leans on: The inverse notationLeans on: Root of a squareLeans on: Absolute valueLeans on: Square of a rootLeans on: Cube and cube rootLeans on: The domain of a square rootLeans on: Clearing small fractionsLeans on: Distributing a negativeLeans on: Domain and range trade placesOpen this lesson
Leans on 1Comes back in 0

1 In plain words

Think of a coat check. You hand over your coat and get a numbered ticket. Later you hand over the ticket and get your coat back. The two trades undo each other: coat to ticket and back to the same coat, or ticket to coat and back to the same ticket. A function and its inverse function behave the same way. Put a number into one, feed the result into the other, and you are holding the number you started with. This lesson shows how to test that with algebra, in both orders, and how it looks on a graph: the two curves are mirror images of each other across the slanted line y = x.

2 Picture it

(0, 3)(3, 0)(1, 4)(4, 1)(2, 7)(7, 2)24682468(0, 3)(3, 0)(1, 4)(4, 1)(2, 7)(7, 2)123456781357
Each labeled point on the solid curve f(x) = x2 + 3, x ≥ 0, has a partner on the dashed inverse x−3 with…

Reminder

Reminder: Reading f(g(x)), The inverse notation, Root of a square, Absolute value, Square of a root, Cube and cube root, The domain of a square root, Clearing small fractions, Distributing a negative, Domain and range trade places
  • Reading f(g(x)). Work from the inside out: g goes first, and f acts on what g gives. With f(x) = x + 1 and g(x) = 2x, f(g(3)) = f(6) = 7.
  • The inverse notation. f⁻¹(x) names the inverse function of f. It does not mean 1 ÷ f(x). For f(x) = x + 4, f⁻¹(x) = x − 4.
  • Root of a square. x2 = |x|, because the root sign gives the answer that is 0 or more: (−5)2 = 25 = 5.
  • Absolute value. |x| = x when x ≥ 0, and |x| = −x when x < 0. For example |4| = 4 and |−4| = 4 = −(−4).
  • Square of a root. (a)2 = a whenever a ≥ 0: (7)2 = 7.
  • Cube and cube root. These undo each other for every real number, with no bars: ∛((−2)3) = ∛(−8) = −2.
  • The domain of a square root. The inside of a square root must be 0 or more: x−3 needs x − 3 ≥ 0, so x ≥ 3.
  • Clearing small fractions. Multiply the top part and the bottom part by the small bottom: (12 + 3) ÷ (12 + 1) becomes (1 + 6) ÷ (1 + 2) = 73 after multiplying both by 2.
  • Distributing a negative. A negative number in front multiplies every term inside: −2(x − 1) = −2x + 2.
  • Domain and range trade places. The outputs of f are the inputs of f⁻¹. If f has domain [0, ∞) and range [3, ∞), then f⁻¹ has domain [3, ∞) and range [0, ∞).

3 Why

The inverse function was built by swapping input and output. So if f sends a to b, the inverse must send b back to a. Run one after the other and a returns as a, which is the algebra test. On a graph, f owns the point (a, b) and the inverse owns (b, a). Swapping the two numbers of a point flips it across the line y = x, because that line holds exactly the points whose two numbers are equal, and those points stay put. Both orders are tested because one order can work while the other fails when a domain was never cut.

4 The rule

Two functions f and g are inverses of each other exactly when g(f(x))=x for every x in the domain of f, and f(g(x))=x for every x in the domain of g.

Then g=f−1, the domain of g is the range of f, the range of g is the domain of f, and the point (a, b) is on the graph of f exactly when (b, a) is on the graph of g. The two graphs are mirror images across the line y=x.

The same idea, five ways

Say it

f of g of x equals x, and g of f of x equals x. When g is the inverse, say: f inverse of f of x is x.

Write it

Doing a function and then its inverse function, in either order, returns every allowed input unchanged, and their graphs are mirror images across the line y = x.

In math
  • f(f⁻¹(x)) = x for every x in the domain of f⁻¹
  • f⁻¹(f(x)) = x for every x in the domain of f
  • (a, b) is on the graph of f exactly when (b, a) is on the graph of f⁻¹
  • domain of f⁻¹ = range of f
  • range of f⁻¹ = domain of f
Like

A coat check: coat to ticket to the same coat, or ticket to coat to the same ticket.

See it
aff⁻¹firstsecondba
Whatever a goes in, f turns it into b and the inverse turns b back into the same a.

The same idea, other ways

As a picture

Draw the slanted line y = x through (0, 0), (1, 1), (2, 2). Take any point of f, such as (2, 7), and swap its two numbers: (7, 2). The two points sit the same distance from the line, on opposite sides, like an object and its reflection. Do this to every point of f and you have drawn the mirror image of f. When f is one-to-one, that mirror image is the graph of the inverse function. If you folded the paper along y = x, the two curves would land on top of each other.

(2, 7)(7, 2)(1, 4)(4, 1)(0, 3)(3, 0)24682468
The line is y = x, and each point above it has a partner below it with the same two numbers in the other…
As a story

You and a friend use a number code. To hide a number, triple it. To read a hidden number, divide it by 3. Hide 5 and you send 15. Your friend divides by 3 and reads 5. It works the other way round too: take 12, divide by 3 to get 4, triple it, and 12 is back. Hiding is f(x) = 3x and reading is f⁻¹(x) = x3. Each one cancels the other, whichever goes first.

5f: multiply by3f⁻¹: divide by3firstsecond155
Hiding turns 5 into 15, and reading turns 15 back into 5.
With small numbers

Let f double a number and add 1. The picture is its table. To find f(3), go down the column under 3: the output is 7. The inverse function asks the opposite question: the output is 7, so which input made it? Find 7 in the bottom row and go up: the input is 3. So f⁻¹(7) = 3, and f⁻¹(f(3)) = 3. The rule for this inverse is f⁻¹(x) = x−12, and (7 − 1) ÷ 2 = 3 agrees with the table.

input xoutput f(x)01132537↑ solve: output given, read every input above it
Reading up from the 7 in the bottom row to the 3 above it is what the inverse function does.
Why it must be true

Say f sends a to b. The inverse function was made to send b back to a. So the inverse of f(a) is a, for every a that f accepts. And f of the inverse of b is f(a), which is b, for every b that f can produce. For the mirror, take (2, 7) and (7, 2). The point halfway between them has x = (2 + 7) ÷ 2 = 4.5 and y = (7 + 2) ÷ 2 = 4.5. Its two numbers are equal, so it sits on y = x. For any pair (a, b) and (b, a) the halfway point is (a+b2, a+b2), always on the line. The line is always exactly in the middle, which is what a mirror does.

As a rule of thumb

Three short checks. One: inside first, and you must come back with x, in both orders. Two: swap the numbers, mirror the point, so (a, b) on f means (b, a) on f⁻¹. Three: domain and range trade places. If any one of the three fails, the pair is not a function and its inverse.

f(f⁻¹(x)) = x

f⁻¹(f(x)) = x

(a, b) on f, so (b, a) on f⁻¹

domain of f⁻¹ = range of f

range of f⁻¹ = domain of f

mirror line: y = x

What to comparef(x)=x2+3,x≥0g(x)=x−3
Steps, in ordersquare, then add 3subtract 3, then take the square root
Domain (allowed inputs)[0, ∞)[3, ∞)
Range (outputs)[3, ∞)[0, ∞)
Where the graph starts(0, 3)(3, 0)
Another point(2, 7)(7, 2)

5 How

  1. Write the domain beside each function, including any cut. The domains decide the last step of the check.
  2. Build f(g(x)): copy the rule of f, and wherever f has x, write the whole rule of g in parentheses.
  3. Tidy up. Squaring undoes a square root, cubing undoes a cube root, and opposite steps such as + 3 and − 3 cancel. The result must be exactly x, with nothing left over.
  4. Build g(f(x)) the same way, with the whole rule of f inside g, and tidy up.
  5. If a square root of a square appears, such as x2, write it as |x| first. Then use the domain to decide whether |x| is x or −x.
  6. If both orders give x, the two functions are inverses. Say where each result holds: f(g(x)) = x on the domain of g, and g(f(x)) = x on the domain of f.
  7. Check with a number. Pick an allowed input a, compute b = f(a), and confirm that g(b) = a. The points (a, b) and (b, a) are mirror images across y = x.

Checking that two functions are inverses

1
Does f(g(x)) tidy up to exactly x for every x in the domain of g?
YesGo on to the other order.
NoStop. They are not inverses. One number that fails to come back is enough to show it.
↓
2
While working out g(f(x)), did a square root of a square show up, such as x2?
YesWrite it as |x|. If the domain says x ≥ 0, then |x| = x. If the domain says x ≤ 0, then |x| = −x.
NoGo on.
↓
3
Does g(f(x)) tidy up to exactly x for every x in the domain of f?
YesThey are inverses on these domains. Finish with a number check and a mirrored pair of points.
NoThey are not inverses as written. Look for a missing cut in the domain or a wrong sign on the root.
  1. Write each function with its domain beside it.
  2. Compute f(g(x)) by putting the whole rule of g, in parentheses, where f has x. Tidy up.
  3. Compute g(f(x)) by putting the whole rule of f, in parentheses, where g has x. Tidy up.
  4. Replace any square root of a square, (something)2, with |something|, and use the domain to remove the bars.
  5. If both orders give exactly x, state that they are inverses and on which inputs each result holds.
  6. Confirm with one number and its mirrored point: f(a) = b should come with g(b) = a.

6 Worked example

Worked example

Checking both orders for a cut parabola and its square root function

Let f(x) = x2 + 3 on the restricted domain x ≥ 0, and let g(x) = x−3. Show that g is the inverse function of f by checking both orders, and name a pair of mirrored points.

Show the steps and the answer
2f: square,then add 3g: subtract 3,then take thesquare rootfirstsecond72
The input 2 goes through f and becomes 7, then goes through g and comes out as 2 again.
  1. Say what the question asks. Feed the output of g into f, then feed the output of f into g. Each time the result must be the x you started with.Inverses must undo each other in both orders. One order alone can pass by accident.
  2. Write the domains. f accepts x≥0, which is the cut. For g, the inside of the root must be 0 or more: x−3≥0, so x≥3.Solving x − 3 ≥ 0 finds the inputs the root accepts. At x = 3 the inside is 0, which is allowed. At x = 2 the inside is −1, which is not.
  3. Build the first order: f(g(x))=(x−3)2+3.f(g(x)) means g works first. The whole rule of g goes where f has x.
  4. Square the root: (x−3)2=x−3.Squaring undoes a square root when the inside is 0 or more, and x ≥ 3 makes x − 3 at least 0.
  5. Finish: x−3+3=x. So f(g(x))=x for every x≥3.−3 and + 3 add to 0. The result holds on the domain of g, the function that worked first.
  6. Build the second order: g(f(x))=(x2+3)−3=x2.Now f works first, so the whole rule of f goes where g has x, and + 3 − 3 cancels under the root.
  7. Write the root of a square with bars: x2=|x|.The root sign always gives the answer that is 0 or more. For example (−5)2 = 25 = 5, not −5.
  8. Use the domain of f. Since x≥0,|x|=x. So g(f(x))=x for every x≥0.The bars change nothing on a number that is already 0 or more. This is the step the cut domain was made for.
  9. Name mirrored points. f(2)=22+3=7, so (2, 7) is on the graph of f. Swap the numbers: (7, 2) is on the graph of g, and g(7)=7−3=4=2 agrees.If f sends a to b, the inverse sends b to a, so every point (a, b) on f has the partner (b, a) on the inverse.
Answer
  • f(g(x)) = x for every x ≥ 3
  • g(f(x)) = x for every x ≥ 0
  • So g = f⁻¹ on these domains. The point (2, 7) on f mirrors to (7, 2) on g.
Double-check Use numbers in both orders. Start with 4: f(4) = 16 + 3 = 19, then g(19) = 16 = 4. Start with 12: g(12) = 9 = 3, then f(3) = 9 + 3 = 12. Both numbers came back. The domains also match: the smallest output of f is f(0) = 3, so the range of f is [3, ∞), which is exactly the domain of g.

Ladder: from easy to exam-hard

Try each rung before you open its steps.

Rung 1 of 7

Add 4, subtract 4

Show that f(x) = x + 4 and g(x) = x − 4 are inverses of each other.

Show the steps and the answer
6f: add 4g: subtract 4firstsecond106
Adding 4 and then subtracting 4 returns the 6 that went in.
  1. Say what is asked: put the result of g into f, then the result of f into g. Each time the answer must be the x you started with. Both functions accept every real number.Inverses must undo each other in both orders, on every allowed input.
  2. f(g(x))=f(x−4)=(x−4)+4.g works first, so the whole rule of g, x − 4, goes where f has x.
  3. (x−4)+4=x.−4 and + 4 add to 0.
  4. g(f(x))=g(x+4)=(x+4)−4=x.Now f works first, so x + 4 goes where g has x, and + 4 and −4 cancel again.
Answer
f(g(x)) = x and g(f(x)) = x for every real x, so f and g are inverses.
Double-check Use numbers. f(6) = 10 and g(10) = 6. In the other order, g(1) = −3 and f(−3) = 1. Each number came back.

Rung 2 of 7

An inverse read from a table and from points

The table shows a one-to-one function f. Find f⁻¹(7), find f⁻¹(f(4)), and name the points on the graph of f⁻¹ that mirror (0, 3) and (3, 9).

input xoutput f(x)03152739411
x: 0, 1, 2, 3, 4; f(x): 3, 5, 7, 9, 11.
Show the steps and the answer
input xoutput f(x)03152739411↓ evaluate: input given, read the output below it↑ solve: output given, read every input above it
Go down a column to use f, and go up from the bottom row to use the inverse.
  1. f−1(7) asks: the output of f is 7, so which input gave it? Find 7 in the bottom row and look up. The input above it is 2, so f−1(7)=2.The inverse function runs the table from the bottom row to the top row.
  2. For f−1(f(4)), work inside first. The column under 4 shows f(4)=11. Then f−1(11) asks which input gives 11, and that input is 4.f carries 4 to 11 and the inverse carries 11 back to 4, so you return to the start.
  3. Swap the two numbers of each point: (0, 3) becomes (3, 0), and (3, 9) becomes (9, 3).If f sends a to b, the inverse sends b to a, so (a, b) on f gives (b, a) on f⁻¹.
Answer
  • f⁻¹(7) = 2
  • f⁻¹(f(4)) = 4
  • (3, 0) and (9, 3) are on the graph of f⁻¹
Double-check Go through the table the forward way. f(2) = 7 agrees with f⁻¹(7) = 2, f(0) = 3 agrees with the point (3, 0), and f(3) = 9 agrees with the point (9, 3).

Rung 3 of 7

A line with two steps

Show that f(x) = 3x − 6 and g(x) = x3 + 2 are inverses.

Show the steps and the answer
  1. Say what is asked: both orders must return x. Both functions accept every real number, so there is no domain to watch.A line that is not flat is a one-to-one function on all real numbers.
  2. f(g(x))=3(x3+2)−6.The whole rule of g goes in parentheses where f has x.
  3. Multiply through: 3·x3+3·2−6=x+6−6.The 3 multiplies both terms inside the parentheses, and 3 times a third of x is x.
  4. x+6−6=x.+ 6 and −6 add to 0.
  5. g(f(x))=3x−63+2.The whole rule of f goes on top of the fraction, where g has x.
  6. Divide each term by 3: x−2+2=x.3x ÷ 3 = x and 6 ÷ 3 = 2, then −2 and + 2 cancel.
Answer
f(g(x)) = x and g(f(x)) = x for every real x, so f and g are inverses.
Double-check Use numbers. f(4) = 12 − 6 = 6 and g(6) = 2 + 2 = 4. In the other order, g(9) = 3 + 2 = 5 and f(5) = 15 − 6 = 9. The points (4, 6) on f and (6, 4) on g are a mirrored pair.

Rung 4 of 7

A cubic and a cube root function

Show that f(x) = 2x3 + 1 and g(x) = ∛(x−12) are inverses. Does either one need a restricted domain?

Show the steps and the answer
(0, 1)(1, 0)(1, 3)(3, 1)(−1, −1)−4−224−4−224
The solid cubic and the dashed cube root trade the numbers of each point.
  1. Write the domains. Both accept every real number.A cube root function takes negative numbers too, and this cubic is a one-to-one function without any cut.
  2. f(g(x))=2((x−12)3)3+1.The whole rule of g goes where f has x.
  3. Cube the cube root: 2·x−12+1.Cubing undoes a cube root for every real number.
  4. 2·x−12=x−1, so the result is (x−1)+1=x.Twice a half of something is the something, and −1 and + 1 cancel.
  5. g(f(x))=((2x3+1)−12)3=(2x32)3=(x3)3.The whole rule of f goes where g has x. Then + 1 − 1 cancels and 2 ÷ 2 = 1.
  6. (x3)3=x for every real x, with no absolute value bars.A cube keeps the sign of its number: (−2)3 = −8 and ∛(−8) = −2. So the cube root hands back the original number, negative or not.
Answer
  • f(g(x)) = x and g(f(x)) = x for every real x
  • They are inverses, and no restricted domain is needed.
Double-check Try a positive and a negative number. f(2) = 2 · 8 + 1 = 17, and g(17) = ∛(162) = 83 = 2. f(−1) = −2 + 1 = −1, and g(−1) = ∛(−22) = ∛(−1) = −1. Both came back.

Rung 5 of 7

The left half of a parabola

Let f(x) = x2 − 4 on the restricted domain x ≤ 0, and let g(x) = −x+4. Show that they are inverses.

Show the steps and the answer
(−3, 5)(5, −3)(−2, 0)(0, −2)(0, −4)(−4, 0)−4−2246−4−2246
Each point of f has a partner on g with its two numbers in the other order.
  1. Write the domains. f accepts x≤0, the left half of the parabola, cut at its vertex (0,−4). For g, the inside of the root must be 0 or more: x+4≥0, so x≥−4.Solving x + 4 ≥ 0 finds the inputs the root accepts. The domain of f will decide what the absolute value becomes.
  2. f(g(x))=(−x+4)2−4.The whole rule of g, minus sign included, goes where f has x.
  3. (−x+4)2=x+4.A negative times a negative is positive, as in (−3)2 = 9, and squaring a square root returns the inside.
  4. x+4−4=x. So f(g(x))=x for every x≥−4.+ 4 and −4 cancel. The result holds on the domain of g.
  5. g(f(x))=−(x2−4)+4=−x2.The whole rule of f goes under the root where g has x, and −4 + 4 cancels.
  6. x2=|x|, so g(f(x))=−|x|.The root sign gives the answer that is 0 or more, which is what the bars say.
  7. On the domain of f, x≤0, so |x|=−x. Then −|x|=−(−x)=x.For a negative number the bars flip the sign: |−3| = 3, which is −(−3). The minus sign in front of the root flips it back.
Answer
  • f(g(x)) = x for every x ≥ −4
  • g(f(x)) = x for every x ≤ 0
  • They are inverses on these domains.
Double-check Use a number from the left half. f(−3) = 9 − 4 = 5, and g(5) = −9 = −3, so −3 came back. With a plus sign on the root the trip would end at 3 instead of −3, which shows why the minus sign belongs to the left half.

Rung 6 of 7

A rational function

Show that f(x) = x+1x−2 and g(x) = 2x+1x−1 are inverses, and state where each check holds.

Show the steps and the answer
5f(x) = (x + 1)÷ (x − 2)g(x) = (2x +1) ÷ (x − 1)firstsecond25
The input 5 becomes 2 in the first machine and comes back as 5 from the second.
  1. Write the domains. For f, x−2=0 at x=2, so f accepts every x≠2. For g, x−1=0 at x=1, so g accepts every x≠1.Setting each bottom equal to 0 finds the one input that would mean dividing by zero, so it can be thrown out. At x = 2 the bottom of f is 2 − 2 = 0.
  2. f(g(x))=(2x+1x−1+1)÷(2x+1x−1−2).The whole rule of g replaces x in the top of f and in the bottom of f.
  3. Multiply the top part and the bottom part by (x−1).Multiplying top and bottom by the same nonzero number does not change a fraction, and it clears the small fractions. Since x ≠ 1, x − 1 is not zero.
  4. Top: (2x+1)+1·(x−1)=2x+1+x−1=3x.The small fraction times (x − 1) leaves 2x + 1, and the 1 becomes x − 1.
  5. Bottom: (2x+1)−2(x−1)=2x+1−2x+2=3.The −2 multiplies both x and −1, and −2 times −1 is + 2.
  6. f(g(x))=3x3=x, for every x≠1.3 divided by 3 is 1.
  7. g(f(x))=(2·x+1x−2+1)÷(x+1x−2−1). Multiply the top part and the bottom part by (x−2).Same move with the other small bottom. Since x ≠ 2, x − 2 is not zero.
  8. Top: 2(x+1)+(x−2)=2x+2+x−2=3x. Bottom: (x+1)−(x−2)=x+1−x+2=3.Subtracting (x − 2) changes both of its signs: −x and + 2.
  9. g(f(x))=3x3=x, for every x≠2.The result holds on the domain of f, the function that worked first.
Answer
  • f(g(x)) = x for every x ≠ 1
  • g(f(x)) = x for every x ≠ 2
  • They are inverses. The domain of f, x ≠ 2, is the range of g, and the domain of g, x ≠ 1, is the range of f.
Double-check Use numbers. f(5) = 63 = 2, and g(2) = 51 = 5. In the other order, g(4) = 93 = 3, and f(3) = 41 = 4. Both came back, and (5, 2) on f mirrors (2, 5) on g.

Rung 7 of 7

Catching a wrong inverse and fixing it

A friend says the inverse of f(x) = (x − 1)2 on the restricted domain x ≥ 1 is g(x) = x − 1. Test the claim. If it fails, fix it and verify the fix.

Show the steps and the answer
  1. Try a number first. Take 4, which is allowed because 4≥1. f(4)=(4−1)2=9. Then g(9)=9−1=3−1=2.If g were the inverse, 4 would come back. One number that fails is enough to show a pair is not inverses.
  2. 2 is not 4, so g is not the inverse of f.The trip did not return to its start.
  3. See the same failure in algebra: f(g(x))=(x−1−1)2=(x−2)2, which is not x. At x=9 it gives (3−2)2=1, not 9.The rule did not reduce to x: at x = 9 it gives 1, so this g does not undo f.
  4. Fix it. f subtracts 1 and then squares. Undo in reverse order: take the square root first, then add 1. The corrected rule is g(x)=x+1 for x≥0.The last step done is the first step undone. The plus root is used because the domain x ≥ 1 is the right half of the parabola.
  5. f(g(x))=(x+1−1)2=(x)2=x, for every x≥0.+ 1 and −1 cancel, and squaring undoes the root when x is 0 or more.
  6. g(f(x))=(x−1)2+1=|x−1|+1.The root of a square is the absolute value of what was squared.
  7. Since x≥1,x−1≥0, so |x−1|=x−1. Then g(f(x))=x−1+1=x, for every x≥1.The bars change nothing on a number that is already 0 or more.
Answer
  • The claim fails: the friend's g sends f(4) = 9 to 2, not back to 4.
  • The correct inverse is f⁻¹(x) = x + 1 for x ≥ 0.
  • f(f⁻¹(x)) = x for every x ≥ 0, and f⁻¹(f(x)) = x for every x ≥ 1.
Double-check Use numbers with the corrected rule. f(4) = 9 and 9 + 1 = 4, so 4 came back. In the other order, 25 + 1 = 6 and f(6) = 52 = 25. The domains match as well: the smallest output of f is f(1) = 0, so the range of f is [0, ∞), the domain of the inverse.

Counterexamples

What it is not, and the tempting wrong moves.

✗ Not this: f(x) = x2 and g(x) = x are inverses, because (x)2 = x.
Only one order was checked. The other order gives g(f(x)) = x2 = |x|, and at x = −3 that is 3, not −3. On all real numbers f is not a one-to-one function: 3 and −3 both give 9, so nothing can send 9 back to both.
✓ Instead: Use the restricted domain x ≥ 0 for f. Then |x| = x, and both orders return x: f(x) = x2 for x ≥ 0 and g(x) = x are inverses.
✗ Not this: f⁻¹(x) means 1 over f(x). For f(x) = 2x the inverse is 12x.
Test it. f(3) = 6, and putting 6 into 12x gives 112, not 3. The small −1 on a function name is a label that means inverse function. It is not an exponent.
✓ Instead: The inverse of f(x) = 2x is f⁻¹(x) = x2. Then f⁻¹(6) = 3, and 3 comes back.
✗ Not this: The inverse of f(x) = 2x + 3 is g(x) = x2 − 3, since dividing undoes multiplying and subtracting undoes adding.
The steps were undone in the same order instead of in reverse. The check shows it: f(g(x)) = 2(x2 − 3) + 3 = x − 6 + 3 = x − 3, which is not x.
✓ Instead: Undo the last step first: subtract 3, then divide by 2. g(x) = x−32. Then f(g(x)) = (x − 3) + 3 = x and g(f(x)) = 2x2 = x.
✗ Not this: To draw the inverse, flip the graph of f over the x-axis.
Flipping over the x-axis sends (2, 7) to (2, −7). That changes the sign of the output and draws −f(x), a different function. Flipping over the y-axis sends (2, 7) to (−2, 7), also wrong.
✓ Instead: Swap the two numbers of every point: (2, 7) becomes (7, 2). The mirror for that swap is the slanted line y = x.
✗ Not this: With f(x) = x3 and g(x) = x, the inputs 0, 1 and −1 all come back, so f and g are inverses.
Those three numbers happen to equal their own cubes. Try 2: f(2) = 8 and g(8) = 8, not 2. A few passing numbers never prove that a pair is inverses.
✓ Instead: Do the algebra: g(f(x)) = x3, which is not x. The inverse of f(x) = x3 is the cube root function f⁻¹(x) = x3, because ∛(x3) = x for every real x.

Tips and tricks

  • Socks and shoes: you put on socks, then shoes. To undo, the shoes come off first. The inverse undoes the steps of f in reverse order.
  • Swap the numbers, mirror the point: (a, b) on f means (b, a) on f⁻¹.
  • A number test can show that a pair is not inverses with one failure. It can never prove that they are. Use the algebra to prove it and a number to catch slips.
  • Whenever (something)2 appears, stop and write bars: |something|. Then let the domain decide the sign.
  • Each result belongs to the function that works first: f(g(x)) = x holds on the domain of g, and g(f(x)) = x holds on the domain of f.
  • Fast domain check: the domain of f⁻¹ must be the range of f. If the smallest output of f is 3, the inverse must start at x = 3.
  • A point on the mirror line, such as (4, 4), is its own mirror image. When you sketch by hand, use the same scale on both axes or the mirror will look crooked.

7 The trap

Checking one order only, or dropping the domain. With f(x) = x2 on all real numbers and g(x) = x, the order f(g(x)) = (x)2 = x works, yet g(f(−3)) = 9 = 3, not −3. Always work both orders, write x2 as |x|, and write the domain beside each result.

Flash cards

Say the answer out loud, then flip.

Which two checks must both come out to x before you can call f and g inverses of each other?
  • f(g(x)) = x for every x in the domain of g.
  • g(f(x)) = x for every x in the domain of f.
  • It is the coat check both ways: coat to ticket and back to the same coat, and ticket to coat and back to the same ticket.
The point (2, 7) is on the graph of f. Which point must be on the graph of f⁻¹, and why?
  • The point (7, 2).
  • The inverse function swaps input and output, so if f sends 2 to 7, the inverse sends 7 back to 2. Swapping the two numbers of a point flips it across the line y = x.
Let f(x) = x2 on all real numbers and g(x) = x. The order f(g(x)) gives x. Why are they still not inverses?
  • The other order fails. g(f(−5)) = 25 = 5, not −5.
  • x2 is |x|, and |x| equals x only when x ≥ 0. One order can work while the other fails when a domain was never cut, so always test both orders.
If g is the inverse function of f, how do the domain and range of g compare with those of f?
  • They trade places.
  • The domain of g is the range of f.
  • The range of g is the domain of f.
  • What f hands out, g takes in, like the tickets a coat check hands out being the same tickets it takes back.

From earlier

Domain and range trade places

f is a one-to-one function. Where do you get the domain of f−1?
Show answer
From the range of f. The backward coat check takes in tickets, and only the tickets that were handed out.

Domain and range trade places

You know f(4) = 9, and f is a one-to-one function. Which number goes into f−1, and which comes out?
Show answer
9 goes in, so 9 is in the domain of f−1.
4 comes out, so 4 is in the range of f−1.
Reason: f(4) = 9 says the same thing as f−1(9) = 4.