Not sure which lesson to do first? Find your level and get a plan →

Math · Advanced Functions · Grade 9-12

Restricted Domain

Restricting a domain deliberately limits which inputs are allowed so a function behaves the way you need — most often to make it one-to-one and therefore invertible.


Quick Answer

Restricting a domain deliberately limits which inputs are allowed so a function behaves the way you need — most often to make it one-to-one and therefore invertible. The formula stays the same; you only keep part of the graph. Restricting f(x)=x2 to x≥0 keeps the right half of the parabola, which passes the horizontal line test and has the inverse x. Use a restriction when a function repeats outputs but you need to undo it, as with sin⁡x on [−π2,π2] for arcsin⁡, or when a model only makes sense for some inputs, such as times t≥0.

Why This Matters

Restrictions are what make inverse functions such as x, arcsin⁡x and arccos⁡x possible; without them, the output 9 of x2 could not be traced back to a single input. They are also why calculators agree: sin⁡−1(0.5) always returns π6, never 5π6, because everyone uses the same interval. Forgetting the restriction leads to inverses that return the wrong branch.

Intuitive Explanation

The function f(x)=x2 sends both 3 and −3 to 9. Handed the output 9 and asked for the input, you have two possible answers, so no single inverse can exist. Cut the domain down to x≥0 and the problem disappears: every output now comes from exactly one input, and the inverse reports that input, 9=3. Nothing about the squaring rule changed; you only erased the left half of the graph.

There is usually more than one good cut. x≤0 also works for x2 and gives the inverse −x. For sin⁡x, any interval where the graph only climbs or only falls would do, but mathematicians agreed on [−π2,π2], where sine climbs from −1 to 1 and takes every output once. For cos⁡x the agreed interval is [0,π]. The restricted domain then becomes the inverse's range, which is why arcsin⁡x always lies between −π2 and π2.

A natural domain is different: the formula forces it on you. x−4 only makes sense for x≥4, and 1x cannot accept 0; nobody chose those limits. A restricted domain is a choice made inside the natural domain, for a purpose. So "find the domain of x−4" asks for the natural domain, while "restrict x2 so it has an inverse" asks you to make a choice.

Think of it like this. Restricting a domain is like closing one of two roads that lead into the same town, so a traveler who arrives can say for certain which road they took. The towns are outputs, the roads are inputs, and closing a road removes inputs without moving any town. Close the wrong roads and an arrival still does not reveal the route.

Core idea. Restricting a domain throws away part of the input set so a function gains a property like invertibility.


When to Use

Restrict the domain when a function fails the horizontal line test but you need an inverse, as with x2, sin⁡x, cos⁡x or ∣x∣, or when a model only applies to some inputs, like a ball's height during the 2.6 seconds it is in the air. Pick an interval on which the function only increases or only decreases, and carry it over as the inverse's range. If a question asks which inputs a formula allows, that is the natural domain; just find where the expression is defined.

Tip. Ask: Am I deliberately limiting the inputs so the function gains a property (like passing the horizontal line test)?

How to Recognize It

Before cutting a function's inputs, ask why the cut is needed: to undo the function, to fit a real situation, or only because the formula breaks somewhere.

  1. Does the formula itself break for some inputs, such as a zero denominator or the square root of a negative?

    If yes, those inputs lie outside the natural domain, which you find rather than choose: x−3 needs x≥3. A restricted domain is an extra cut made on purpose, even where the formula would still work.

  2. Will you need to undo the function, and does some output come from two different inputs?

    If yes, keep an interval where the graph only rises or only falls, such as x≥0 for x2 or [0,π] for cos⁡x. If every output already has one input, as with 2x+5, invert at once.

  3. Does the situation make sense only for some inputs, such as times before a ball lands?

    If yes, the restriction comes from the context: h(t)=−4.9t2+12t+1 models the ball only from t=0 until it lands, about 2.53 seconds later. The algebra accepts negative times, but the story does not.

  4. Have you found an inverse formula but not yet said which inputs it accepts?

    If yes, carry the restriction across: the restricted domain of f becomes the range of f−1. Restricting (x−2)2 to x≥2 gives f−1(x)=2+x for x≥0, whose outputs are all at least 2.

Restricted Domain compared with Domain (natural), Horizontal line test, and Inverse function

These four rows are steps in one chain, which is why they blur: test whether outputs repeat, cut the inputs if they do, then build the inverse. Ask whether the set of inputs is forced by the formula or chosen by you for a purpose.

Restricted Domain

Meaning
Use this when a function lacks a needed property (usually invertibility) on its full domain and you cut the inputs down to an interval where the property holds.
Key test
Am I deliberately limiting the inputs so the function gains a property (like passing the horizontal line test)?
Example
On what domain can you restrict f(x)=x2 so it has an inverse, and what is that inverse?

Domain (natural)

Meaning
The full set of inputs for which the rule is defined, before any deliberate cutting.
Key test
Use when finding where a function is simply defined (no division by zero, no negative roots).
Example
Domain of 1/x is x≠0

Horizontal line test

Meaning
Diagnoses WHETHER a function is one-to-one; restriction is the FIX when it fails.
Key test
Use to check invertibility; then restrict if it fails.
Example
x2 fails until restricted

Inverse function

Meaning
The reversal you obtain AFTER restricting; restriction is the prerequisite step.
Key test
Use once the function is one-to-one and you want to swap inputs and outputs.
Formula
f−1
Example
x inverts x2 on x≥0

Formula & Notation

How to read it: Example: “f:[0,∞)→[0,∞)”.

Worked Examples

Example 1 — Restrict to make invertible

Easy

On what domain can you restrict f(x)=x2 so it has an inverse, and what is that inverse?
Show the steps and the answer
  1. Show the problem: f(3)=9 and f(−3)=9.

    Two inputs share an output, so an inverse would not know which to return.

  2. Keep only x≥0, where x2 is increasing.

    On an interval where the function only climbs, each output comes from exactly one input.

  3. Invert: solving y=x2 with x≥0 gives x=y, so f−1(x)=x.

    The restriction selects the nonnegative root, which is exactly what   returns.

  4. Check: f−1(f(3))=9=3 and f(f−1(5))=(5)2=5.

    The two functions undo each other on the restricted domain.

Answer: Restrict to x≥0; inverse is x

Takeaway: Cut to an interval where the function only increases or only decreases; x≤0 would also work, giving the inverse −x.

Example 2 — Looks like restriction but is the natural domain

Standard

What is the domain of f(x)=x−4?
Show the steps and the answer
  1. Find what the formula requires: the expression under the square root must be at least 0.

    A negative number has no real square root.

  2. Solve x−4≥0: x≥4.

    This is the largest set of inputs the rule accepts, its natural domain.

  3. Note that nothing was chosen: inputs below 4 are excluded by the formula itself.

    A restriction would be an extra, deliberate cut inside this set, such as 4≤x≤8 for a model.

  4. Check the edge: f(4)=0=0 is defined, while f(3)=−1 is not real.

    The boundary point 4 belongs to the domain.

Answer: x≥4

Takeaway: The natural domain is forced by the formula; a restricted domain is chosen inside it, for a purpose.

Example 3 — Find the mistake: a cut that still repeats outputs

Application

To define an inverse of f(x)=cos⁡x, Ava restricts it to [−π,π] and writes: "Now cos⁡−1(0.5) is the input in [−π,π] whose cosine is 0.5." Find the mistake and correct it.
Show the steps and the answer
  1. Test her interval: cos⁡π3=0.5 and cos⁡(−π3)=0.5, and both inputs lie in [−π,π].

    Two inputs still share an output, so her inverse would not know which to return.

  2. Find why: on [−π,π] cosine rises from −1 to 1 and then falls back to −1.

    A function that turns around inside the interval fails the horizontal line test there.

  3. Choose [0,π], where cosine falls steadily from 1 to −1.

    A strictly decreasing function takes each output once, and this interval still reaches every output from −1 to 1.

  4. Evaluate: cos⁡−1(0.5)=π3, the only input in [0,π] with cosine 0.5.

    The restricted domain [0,π] becomes the inverse's range.

Answer: Restrict to [0,π]; then cos⁡−1(0.5)=π3

Takeaway: A restriction must make the function strictly increasing or strictly decreasing, not just shorter.

Where You'll See It

  • Finding a ladder's angle

    A 5 m ladder reaching 2.5 m up a wall makes an angle with sin⁡θ=0.5. A calculator returns sin⁡−1(0.5)=30∘, never 150∘, because arcsin⁡ uses the restricted domain from −90∘ to 90∘, which fits a ladder leaning against a wall.

  • When a thrown ball reaches 6 m

    A ball's height is h(t)=1.5+12t−4.9t2 m, and the model only applies while it is in the air, 0≤t≤2.57 s. It reaches 6 m twice, at t≈0.46 s and t≈1.99 s. Restricting to the way up, 0≤t≤1.22 s, gives each height a single time.

Common Mistakes

  1. The slip: Thinking that restricting the domain changes the function itself.

    Why it's tempting: Restriction does not rewrite the formula. f(x)=x2 restricted to [0,∞) still squares its input — the rule is identical. What's changed is which inputs we allow. Outside the restriction, we just say "that input isn't part of this function anymore."

    Instead: Same formula, fewer allowed inputs. The graph looks like the original with the unwanted half erased — not redrawn.

  2. The slip: Restricting the range (outputs) instead of the domain (inputs).

    Why it's tempting: Inverting needs each output to come from exactly one input. So we limit the inputs. Limiting outputs doesn't fix the underlying problem — multiple inputs can still map to the same surviving output.

    Instead: To create x from x2, restrict the domain to [0,∞). The range adjusts on its own.

  3. The slip: Confusing valid x-values with valid y-values when defining the inverse.

    Why it's tempting: When you swap x and y to find an inverse, the original range becomes the new domain, and the original restricted domain becomes the new range. Students lose track of which set is which.

    Instead: Anchor: f−1 undoes f. So f−1's inputs are whatever f used to output. Example: f(x)=x2 on [0,∞) outputs [0,∞). So f−1(x)=x is defined for x∈[0,∞).

  4. The slip: Picking a restriction interval where the function is still not one-to-one.

    Why it's tempting: If you restrict cos⁡(x) to [−π,π], cosine is still not monotonic on that interval — it goes down, then back up. The restricted function still fails the horizontal-line test, and the inverse still doesn't exist.

    Instead: Strictly monotonic — not just "mostly." For cosine, the standard restriction is [0,π] — strictly decreasing the whole way.

  5. The slip: Forgetting that the range becomes the domain of the inverse.

    Why it's tempting: Students restrict f, find the inverse formula, and then declare the inverse is defined on all of R — forgetting that the inverse can only accept inputs that the original f could output.

    Instead: Always state both the restricted domain and the resulting range. The inverse swaps them. If f:[0,∞)→[0,∞), then f−1:[0,∞)→[0,∞).

  6. The slip: Assuming there's exactly one "right" restriction.

    Why it's tempting: For x2, both [0,∞) and (−∞,0] produce valid one-to-one restrictions and valid inverses (positive square root vs. negative square root).

    Instead: Multiple restrictions usually work. Conventions ("principal value") pick one — but it's a choice, not a necessity. Calculators all agree because they all use the same convention.

  7. The slip: Conflating natural domain with restricted domain.

    Why it's tempting: The natural domain is the largest set where the formula is defined (e.g., x's natural domain is [0,∞) because the square root needs non-negative inputs). A restricted domain is a deliberately smaller subset, chosen to enforce some property like invertibility.

    Instead: Natural domain = where the formula is even defined. Restricted domain = a chosen subset of that, smaller, with a purpose.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. Restrict f(x)=(x−3)2 to its right half so it has an inverse. Give the restricted domain and the inverse.

    Hint: Cut the parabola at its vertex.

  2. f(x)=∣x+1∣ fails the horizontal line test. Give a restriction that makes it one-to-one, and the inverse on that piece.

    Hint: Where is the corner of the V?

  3. sin⁡x is restricted to [−π2,π2] to define arcsin⁡. What is the range of arcsin⁡, and does arcsin⁡(sin⁡π) equal π?

    Hint: The domain you keep for sin⁡ becomes the output set of its inverse.

  4. A ball's height is h(t)=−5t2+20t metres, t seconds after it is thrown from the ground. What domain makes sense?

    Hint: When is the height zero?

  5. What is the domain of g(x)=1x−5? Is it a natural domain or a restricted one?

    Hint: Is the limit forced by the rule itself, or chosen on purpose?

  6. A student writes: "Restrict f(x)=x2−4x to x≥0, and it has an inverse." Correct the mistake.

    Hint: Find the vertex, then compare f(0) and f(4).

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

Why is 9 equal to 3 and not −3, when both square to 9?

Because x\sqrt{x} is defined as the inverse of x2x^2 restricted to x≥0x\ge 0, so it returns only the nonnegative answer. That is a convention, not a law; restricting to x≤0x\le 0 instead would produce the function −x-\sqrt{x}. Solving x2=9x^2=9 is a different question, which inputs give 9, so its answer is x=±3x=\pm 3.

Why does arcsin⁡(sin⁡5π6) come out as π6?

Because arcsin can only return angles in its restricted range, [−π2,π2]\left[-\frac{\pi}{2},\frac{\pi}{2}\right]. sin⁡5π6=12\sin\frac{5\pi}{6}=\frac{1}{2}, and the only angle in that range with sine 12\frac{1}{2} is π6\frac{\pi}{6}. So arcsin⁡(sin⁡x)=x\arcsin(\sin x)=x holds only when xx already lies inside the restriction. Calculators follow the same rule, which is why they report one angle and leave the other solutions to you.

How do I draw a restricted function in Desmos or on a graphing calculator?

In Desmos, put the condition in braces after the formula, as in y=x2 {x≥0}y=x^2\,\{x\ge 0\}, and only the right half of the parabola is drawn. On many graphing calculators you can divide by a true-or-false test, as in x2(x≥0)\frac{x^2}{(x\ge 0)}, which is undefined wherever the condition is false, so those points are not plotted. Graphing the restricted piece is a quick check that it passes the horizontal line test.

Can a restricted domain be made of two separate pieces?

Yes, as long as no output repeats across the pieces. For x2x^2, keeping [1,2][1,2] and [−4,−3][-4,-3] works, because the outputs [1,4][1,4] and [9,16][9,16] do not overlap. Keeping [1,2][1,2] with [−2,−1][-2,-1] fails, since both pieces produce [1,4][1,4]. Most textbooks still prefer a single interval, because the inverse is then one tidy formula.

Learning Path

Restricted Domain

You are here

Next →

Nothing after this yet
You need Domain (which inputs a formula allows), Function, and Inverse Function (swapping inputs and outputs). This topic adds the deliberate cut that turns a many-to-one function into a one-to-one function. It is the step behind the inverse trigonometric functions arcsin⁡, arccos⁡ and arctan⁡, behind x as the inverse of squaring, and behind piecewise models that apply a rule only on part of the number line.

See Also