The range of a function given as ordered pairs is the set of all second coordinates, with duplicates removed. List every y-value that appears in the pairs, delete repeats, and sort the result; for the relation {(2, 5), (3, 5), (4, 9)}, the range is {5, 9}. That one rule covers almost every textbook and test question β€” the complications come from deciding whether a relation counts as a function at all, and from what to do when the pairs are reversed.

Most explanations of this topic stop at "collect the y-values." That's fine for a single drill problem. It leaves you stuck the moment a question asks whether the relation is a function, swaps the coordinates, or expects interval notation instead of a roster. The method below handles those cases, starting with the coordinate check that has to come first.

Sort Each Pair Into x and y Before Writing Anything Down

An ordered pair is written (x, y), and the order is not negotiable. In (7, 2), 7 is the input and 2 is the output. Say the pair is for a function f; then f(7) = 2, which makes 2 a range value and 7 a domain value. Swapping them changes the answer to a different question.

The x-coordinates collect into the domain, the y-coordinates into the range. The domain tells you what inputs the relation accepts; the range tells you what outputs it produces. For a set like {(1, 3), (2, 3), (5, 8)}, the domain is {1, 2, 5} and the range is {3, 8}. Two different inputs can share an output without any problem, which is why the repeated 3 collapses to one entry.

Pick a consistent notation early because instructors grade on it. Curly braces around the values are standard: {3, 8}. Interval notation, [3, 8], is only correct here if the relation genuinely fills every value in between, and three isolated points do not. That distinction trips up students who learn interval form for graphs and then apply it to discrete pairs.

One practical habit pays off across every problem type. Before listing anything, rewrite each pair vertically as x above y, or mark the y-value in a second column of a scratch table. Mistakes in this topic are almost never about the concept; they're about grabbing the first number of the pair out of habit, especially when you're working fast under exam time pressure.

The Core Method: Collect the Second Coordinates and Remove Repeats

A relation is just a set of ordered pairs, and its range is the set of second elements. The word "set" does real work here. Sets don't count duplicates, so if the y-value 4 shows up four times across the pairs, the range lists 4 once. Students who write {4, 4, 4, 4} get marked down despite doing the hard part correctly.

Lists are order-independent. A range written as {9, 5, 2} is the same set as {2, 5, 9}, and some style guides prefer ascending order for readability. Descending order is not wrong unless your course specifies a convention. Check your textbook's answer key format once, then match it for the rest of the term rather than guessing each time.

Here's the procedure reduced to its bones:

1: Read each ordered pair and isolate the second coordinate. 2: Write the y-values as they appear, in a row or column, including repeats. 3: Cross out every duplicate, keeping one copy of each value. 4: Reorder the survivors ascending if your course prefers it. 5: Wrap the final list in curly braces.

Negative values need the same care as positive ones. For {(-2, 1), (0, -3), (5, 1), (7, 0)}, the second coordinates are 1, -3, 1, 0. Remove the duplicate 1. The range is {-3, 0, 1}. Dropping a minus sign is one of the two most common errors in the whole procedure, alongside reading the x-column by mistake.

Integers aren't the only possible values, so don't round. Fractions like 1/2 and decimals like 0.75 belong in the range exactly as written. So do irrational values such as √2 when a pair contains one. You're copying outputs, not simplifying them, and a simplified version that doesn't match the pair is a wrong answer even if the numbers are close.

Removing Duplicates Actually Belongs to the Domain Check

Here's where most walkthroughs blur two ideas. The range list is a set, so duplicates drop for that reason alone. A second, separate test asks whether the relation is a function at all: no x-value may repeat with a different y. Those are different rules with different consequences, and the domain side is the one that can disqualify the relation.

Check the domain first when a problem mentions functions. Take {(1, 2), (1, 5), (3, 7)}. The input 1 produces both 2 and 5, so the relation fails the function test. This relation still has a range β€” {2, 5, 7} β€” but it isn't a function, and many questions ask you to say so explicitly. Reporting the range without flagging the failure costs points on multi-part questions.

Repeated inputs paired with the same output are fine. The relation {(4, 6), (4, 6), (4, 6)} is a function because every instance of the input 4 yields 6. It looks strange, and some teachers treat a pair written twice as a typo rather than a distinct element, but the set reduces to a single pair anyway: domain {4}, range {6}.

Use the table below to separate the two tests, since they get conflated constantly:

Check What you compare Rule Example on {(1, 2), (1, 5), (3, 7)} Result
Range All y-values Duplicates removed because it's a set {2, 5, 7} Valid range
Function test x-values against their y-values Repeated x with different y = not a function x = 1 gives 2 and 5 Fails
Repeated output Same y on different x Always allowed (2, 9), (4, 9) Still a function

A vertical-line reading helps if you sketch the points. Any repeated x with different y values stacks two points on the same vertical line, which is exactly the failure a vertical line test detects on a graph.

Reading Reversed Pairs and "Find the Domain" Questions Without Mixing Them Up

Sometimes a problem gives pairs as (y, x) or as "input, output" with the labels flipped. If a question states the first coordinate is the output, then the range category moves to the first coordinate. Don't apply the default (x, y) convention when the problem has redefined it. Two seconds spent reading the labels prevents an answer that's fully inverted.

A related trap: "find the domain" questions feed the same pairs through the same logic but keep the first coordinates. The two procedures are mirrors, so if you've already sorted the pairs into columns, both answers come from one table. Skipping that sorting step is what produces the mix-up in the first place.

Watch for function notation folded into the same exercise. A pair like (3, f(3)) with f(3) = -1 becomes (3, -1), and the range value is -1. Composition, where a pair's output feeds another function's input, sits beyond this method's scope and rarely appears in pair-listing exercises, but recognizing when an exercise has crossed that line saves you from forcing the wrong technique.

Textbook conventions vary at the margins: some write ranges as inequalities or intervals, and a few online homework systems auto-grade order. None of these change the core answer. They change the presentation, and matching presentation is part of getting the problem right on an automated grader.

Scatter the Points When You Need to See Whether the Range Is Continuous

Pairs can describe a handful of isolated points or a sample taken from a continuous curve. With four or five pairs, the roster is the honest answer because you have no evidence about what lies between them. With a dozen pairs following an obvious pattern, the problem usually wants you to infer the underlying rule and report a range over an interval instead.

Suppose a set of pairs traces a parabola opening upward, with a vertex output of -4 and both arms rising. The pairs themselves may show only a few points, but the function they sample has a range of [-4, ∞). Reporting {-4, -1, 3, 8} would describe your sample, not the function. Read the question's wording: "the relation shown" means roster; "the function shown" often means infer the full behavior.

Plotting makes the call easier. Sketched points that form a straight diagonal line suggest all real values along that path, while scattered points that follow no pattern keep the roster. This is a judgment call, and it's the one place in this topic where two careful students can land on different answers legitimately, depending on the problem's framing.

For tables in a data-science or algebra context, the same logic applies with heavier vocabulary. A frequency table of inputs and outputs has a range equal to the set of distinct output cells, regardless of how many rows share a value. According to the OpenStax Algebra and Trigonometry (2nd edition) treatment of relations and functions in its "Functions and Function Notation" section, the range is defined exactly this way, as the set of second components, with a warning that repeated values appear only once.

Comparison: Tally the Distinct Outputs, Then Choose Roser or Interval

Two methods get you to the same answer for plain pair lists, and the table below lays out where each one earns its place:

Method How it works Best for Notation Weak point
Tally-first List all y-values with repeats, then cross out duplicates Any pair list; easy to audit under exam pressure Set braces Clutter with long lists
Sorted-distinct Keep a running list of new y-values, adding only unseen values Larger data sets and spreadsheets Set braces Easy to skip a value while scanning
Infer-the-function Fit the pattern and report an interval Pairs sampled from a smooth curve Interval or inequality Wrong when the points are genuinely discrete
Software tally Run the list through a distinct-value count 20+ pairs; verifying homework Whatever the tool outputs Can't judge whether the answer should be an interval

For anything longer than about ten pairs, the tally-first method turns error-prone. Keep a running list of distinct values as you scan; the collection never grows past the number of unique outputs, so it stays short. This is the method to use when a graded quiz hands you twenty or thirty pairs.

The tiebreaker between roster and interval is evidence, not preference. If the pairs are isolated exact values with no stated pattern, the roster is correct and an interval overstates what you know. Software that drops the y-values into a column and counts distinct entries is a legitimate check on your work, but it inherits your framing: hand it discrete points and it returns a roster, not an interval.

Worked Example and the Failure Cases That Cost Points

Run a messy set all the way through: {(0, 4), (2, -1), (2, 4), (5, 4), (7, -1), (9, 0)}. Reading second coordinates gives 4, -1, 4, 4, -1, 0. Removing duplicates leaves 4, -1, 0. Ascending order: {-1, 0, 4}. The domain is {0, 2, 5, 7, 9}, and since the input 2 appears with outputs -1 and 4, the relation fails the function test even though the range is well defined.

Now the failures. First failure: a student reports the range as {-1, 0, 0, 4} by missing a duplicate, which automatic graders mark wrong on the set-equality check even though the values are the right ones. Second: a student reports the domain instead of the range and is fully inverted. Third: a student writes the range in interval form as [-1, 4] on a discrete list, which asserts values like 1.5 that no pair produces.

An edge case with empty output lists doesn't arise in practice, but there's a subtler empty case worth recognizing. A question that says "find the range of the inverse" gives you the original domain, not the original range, because inversion swaps the coordinates. Swap first, then read; don't read first and forget to swap.

A self-check catches most of this. After writing your answer, verify each listed value appears as some pair's second coordinate, verify no duplicate survives, and verify you haven't added a value that isn't in the pairs. Three passes, under a minute, and they catch the errors that actually show up on graded work.

One Pair List, Several Answers: Domain, Range, Function Status, and Inverse

On multi-part test questions, the same pair set produces four answers, and only one of them is the range. Organize the scratch work as a single table with x and y columns; then the domain reads down the first column, the range down the second, the function test compares repeated x rows, and the inverse's range is just the original domain relabeled. One setup, four answers.

Don't confuse range and inverse. The inverse relation flips every pair, so the original range becomes the inverse's domain and the original domain becomes its range. If both the relation and its inverse pass the function test, the function is one-to-one, which every horizontal line test on its graph confirms. That bigger property is beyond what a pair list alone needs to establish, but the coordinate swap is not.

Notation discipline is worth a final mention because it's where correct answers lose points. Sets take braces; ascending order is conventional; decimals and negatives are copied, not rounded; interval notation waits until the question says the values are continuous. None of these are cosmetic details on automated grading systems, which compare answer strings.

Pair set Domain Range Function?
{(2, 5), (3, 5), (4, 9)} {2, 3, 4} {5, 9} Yes
{(1, 2), (1, 5), (3, 7)} {1, 3} {2, 5, 7} No
{(4, 6), (4, 6)} {4} {6} Yes
{(0, 4), (2, -1), (2, 4), (5, 4), (7, -1), (9, 0)} {0, 2, 5, 7, 9} {-1, 0, 4} No

Reading the row for the fourth set is the whole lesson in miniature: the range takes the time the domain check does not, because removing duplicates and testing for repeated inputs are different jobs done on different columns with different consequences.

FAQ

Can a range value repeat in the answer?

No. The range is a set, so each value appears once no matter how many pairs produce it. A relation with 500 pairs whose outputs are all 7 has a range of {7}. If your answer lists a value twice, you transcribed without collapsing, and automated graders comparing sets will reject it.

What's the difference between the domain and the range of a relation?

The domain is the set of first coordinates, the range the set of second coordinates. On {(1, 3), (2, 3), (5, 8)}, the domain is {1, 2, 5} and the range is {3, 8}. Same pairs, different columns, different answers, and both are frequently asked for together.

Is a relation with a repeated output still a function?

Yes, as long as repeated inputs keep the same output. The failure condition is a repeated x with different y values, because that input produces two outputs. Take {(2, 9), (4, 9)}: the output 9 repeats across different inputs, and that relation is a perfectly valid function.

What is the range of an inverse relation built from the same pairs?

It's the original relation's domain. Inversion swaps every coordinate, so first coordinates become second coordinates and vice versa. For {(2, 5), (3, 7)}, the inverse is {(5, 2), (7, 3)}, and its range is {2, 3}, which matches the original domain.

How do I write the range if the pairs come from a continuous curve rather than isolated points?

Use interval or inequality notation over the values the function actually takes. The roster method describes only the sampled points, so when the exercise says the function is continuous, look for a vertex or endpoint value in the pairs and express the range from there, upward or downward as the shape dictates.

Which is the single most common mistake on this type of question?

Reading the first coordinate by habit. It's quick, it feels natural at speed, and it produces an answer that's precisely inverted. Build the two-column table before extracting anything; the column headings do the discipline the clock won't let you do.

Does the range change if the same pair appears twice in the list?

No. Duplicate pairs vanish when the relation is treated as a set, so the domain and range come out identical, and the function status doesn't change. Some instructors treat a repeated pair as a transcription error and ask you to confirm the source list before scoring.