You read a number line into interval notation by taking the leftmost point first, the rightmost point second, and choosing a parenthesis or bracket for each end based on whether that endpoint is filled or open. A filled dot at 3 with an open dot at 7 becomes [3, 7). If the shading runs off the end of the line, that side becomes negative or positive infinity, which always takes a parenthesis. That three-part habit β€” order, endpoint type, infinity rule β€” handles almost every graph you will meet.

Step 1: Scan the Number Line From Left to Right Before Writing Anything

Start every conversion with a silent left-to-right scan. Your eye should sweep across the entire line and note the lowest shaded value, the highest shaded value, and the status of each dot at the ends of the shaded stretch. Writing before scanning causes the single most common error: reading endpoints with open and closed statuses swapped.

Establish the scale first. If the ticks are labeled every 2 units, a dot sitting halfway between labels is at an odd value, not the nearest even one. A graph with tick marks but no visible numbers is unreadable β€” ask for the scale or verify it against a labeled axis in the same problem set.

Watch for gaps and islands. One number line can carry two or three separate shaded stretches, and each stretch is its own interval. Students routinely merge them into one interval by reading only the outermost endpoints, which produces an answer that describes values the graph never shaded.

Number the endpoints in your head as you find them. Say "left end is negative two, open" and "right end is five, closed" before touching paper. The verbal pass forces you to commit to endpoint type, which is the information the bracket or parenthesis must carry.

Step 2: Decide Filled Dot vs Open Dot and Match It to a Bracket

A filled, solid dot means the endpoint is included in the set. An open, hollow dot means it is excluded β€” the value is a boundary you approach but never reach. The bracket is the transcription of that dot, which is why misreading the dot guarantees a wrong interval even when the numbers are correct.

Endpoint marking Meaning Bracket to write Set-builder equivalent
Filled dot at a a is included [ or ] x β‰₯ a or x ≀ a
Open dot at a a is excluded ( or ) x > a or x < a
Arrow or line running off the axis unbounded ( or ) always x > a, x < a, or all reals
Two dots of any kind bounded interval one of () [] ][ between the two values

Match the bracket's direction too. The bracket or parenthesis at the left end opens toward the right, and the one at the right end opens toward the left, so [3, 7) puts the square bracket on the 3 and the parenthesis on the 7. In many U.S. textbooks the convention is simply that the square bracket "hugs" the number it includes.

The one exception worth memorizing is reversed notation, written [a, b) with the bracket facing backward. Some textbooks use it to signal that the endpoint is excluded, which flips the usual meaning. If your course uses that convention, confirm it in the textbook's notation appendix before the exam rather than assuming the standard reading.

Step 3: Convert Rays and the Whole Line Using Infinity Parentheses

Shading that continues past the last labeled tick with an arrowhead is unbounded on that side. That side always gets a parenthesis, regardless of the dot at the other end. The symbols are βˆ’βˆž for the left and +∞ (usually written just ∞) for the right, and infinity is never a value you can reach, so a bracket there is always wrong.

Graph appearance Interval notation Read aloud
Open dot at 2, shading right (2, ∞) all reals greater than 2
Filled dot at βˆ’5, shading left (βˆ’βˆž, βˆ’5] all reals at most βˆ’5
No dots, shading across everything (βˆ’βˆž, ∞) all real numbers
Open dot at 0, shading right (0, ∞) positive reals only

The whole line gets the same treatment as a ray. When a single line shades every point from edge to edge with arrowheads at both ends, the interval is (βˆ’βˆž, ∞), which is the full set of real numbers. One popular misconception says a filled dot at zero alongside a full-line shade should become [0, ∞).

Beware the open dot with no shading visible beyond it. If a graph shows a hollow dot at 4 and the line stops rather than continuing, that is a degenerate interval with no width, not a set. Check with your instructor whether the problem intends a single excluded point or a ray, because the drawing alone is ambiguous.

Step 4: Handle Double-Shaded Graphs by Writing a Union

Two separate shaded stretches on one axis demand a union, written with the cup symbol βˆͺ between the intervals. Read each stretch independently, fully, and then connect them. For example, shading from βˆ’3 (open) to 1 (filled) plus shading from 4 (filled) rightward becomes (βˆ’3, 1] βˆͺ [4, ∞).

Decide the direction of the union before writing: left interval first, then the right. Reading order is not mandatory mathematically, since union is commutative, but it prevents dropped pieces and makes grading easier.

Watch the spacing and the bracket count. A union answer has two intervals, so it needs two pairs of brackets or parentheses, four characters total. Answers with three symbols usually mean one end was forgotten.

If the two stretches touch at one filled point, the union can collapse into a single interval. Shading from 0 (open) to 2 (filled) plus 2 (filled) rightward is just (0, ∞), because the point 2 is already included in the second ray and the first stretch reaches toward it from below without a gap. Confirm by asking whether any real number between the outermost bounds is missing from both stretches.

Step 5: Translate Word Phrases and Inequalities Into the Same Brackets

Many problems give the interval in words or inequalities and expect the notation. The mapping is mechanical once the endpoint rule is settled. "x is at least 4 and less than 9" is [4, 9); "x is greater than βˆ’1 or at most 6" is (βˆ’1, 6]; "x is between 0 and 2, inclusive" is [0, 2]. The phrase "between" is the trap, since it can be exclusive in plain English but inclusive in some textbooks.

Phrase Typical notation Endpoint rule applied
at least 4, less than 9 [4, 9) at least = included, less than = excluded
greater than βˆ’1, at most 6 (βˆ’1, 6] greater than = excluded, at most = included
between 0 and 2, inclusive [0, 2] "inclusive" forces both brackets
no more than 10 (βˆ’βˆž, 10] no lower bound, upper included
any real number (βˆ’βˆž, ∞) no bounds at all

Inequality chains convert the same way. The compound statement βˆ’2 ≀ x < 5 becomes [βˆ’2, 5) with the bracket on the "or equal to" side. A strict chain like 1 < x < 3 becomes (1, 3), with parentheses on both sides. If the directions conflict, split the compound into separate pieces or rewrite it as a union.

Converting from interval notation back to a number line is the same process in reverse and makes a good self-check. Draw the dots at the endpoint values, fill them where the bracket is square, open them where it is round, and shade between for bounded intervals or outward for infinite ones. When your redrawn graph matches the original, the notation is consistent.

Step 6: Check Bracket Direction and Run a Number Test

Take the finished interval and test a value that the graph does contain, then a value it does not. For [3, 7), the number 6 should satisfy the inequality 3 ≀ x < 7, and the number 7 should not. If the inequality you wrote from the notation admits 7, the parenthesis at the right end is in the wrong place.

Test the boundary numbers themselves, not just interior values. Use the exact endpoint and ask whether it counts. Confusion tends to appear at the included end of a mixed interval, where people write [3, 7] but intended to exclude 7.

Verify the bracket direction visually. On the left, a square bracket opens right toward the interval's interior. On the right, it opens left. Reversed brackets can pass a casual glance and still fail a grader, so look at each character individually.

Run one last check against the original problem's conventions. Some software platforms, including graphing tools built on Desmos, accept bracket notation directly in expressions, while others require inequalities; the typed notation and the displayed notation can differ. If you are entering an answer into an online system, read the platform's help page for its accepted interval format, since a correct interval entered with the wrong syntax still gets marked wrong.

Numbers That Get Misread on a Sparse Axis

Scale errors are harder to fix than bracket errors because they look like arithmetic mistakes. Before converting, confirm the step between tick marks. Common spacings are 1, 2, 0.5, and 10 units, and a dot exactly on an unlabeled tick belongs to that tick's value once the step is known. Halfway positions require a smaller step.

Zero is the calibration point. Find where the axis crosses zero before estimating any other position. If the graph shows only positive numbers and the shaded region approaches the y-axis, the endpoint might be 0, not 1. For example, a shaded region ending at the axes should be tested for x β‰₯ 0 first, especially if the problem came from a context like domain restrictions.

Negative endpoints are read by distance from zero, not by tick count from the left. A dot three ticks left of zero is βˆ’3 only if the step is 1. Confirm the step by counting the ticks between two labeled numbers and dividing.

A worked example with a sparse axis: ticks labeled βˆ’10, βˆ’5, blank, 5, 10, with a filled dot two ticks left of the 5 label and an open dot at the βˆ’5 label. The step is 5; the filled dot sits at βˆ’5? No β€” count from 5 leftward by 5 per tick: 0, then βˆ’5. The two ticks left of 5 are 0 and βˆ’5, so the endpoint is 0. The interval is (βˆ’5, 0], with a parenthesis at the open βˆ’5 and a bracket at the filled 0.

Reading Unbounded Endpoints Without the Infinity Symbol

Graphs that shade to the edge of the paper usually mean "all the way." A dot and then an arrow, shading all the way to the border of a printed grid, is the printed equivalent of unbounded. The infinity symbol is a notation convenience; it is not a value, which is why it never gets a bracket.

Some teachers shade a whole region using a thick arrow beyond the dot. Treat that as the same thing as an arrow o across many lines. A solid shading to the left edge with no arrowhead is usually still intended as unbounded in a textbook diagram, but if the problem wording says "shown in the graph only," check with the instructor.

An infinity in the wrong slot is easy to catch. If a filled dot appears at the leftmost position and no shading reaches the left edge, then the interval is bounded on the left and infinite on the right only. The correct notation is [2, ∞), never (βˆ’βˆž, 2], which describes the opposite ray.

One comparison to keep in mind: a horizontal line graph (a constant function) has a full-line domain, (βˆ’βˆž, ∞), but its range is a single value, often written as a set {k} or as a degenerate interval [k, k]. Interval notation can express [k, k] but not a single-point set in the usual union form, and many textbooks ban degenerate intervals entirely. Ask which convention your course uses before writing one.

Worked Conversions: Three Graphs, Three Intervals

The fastest way to lock in the method is to run three contrasting examples back to back. This first shaded step, a filled dot at βˆ’1 with shading rightward, gives [βˆ’1, ∞), and the square bracket is doing the work of the fill. If that dot had been hollow, the same shape would become (βˆ’1, ∞).

The second example is a bounded pair: an open dot at 2 and a filled dot at 8, shaded between. That is (2, 8], since the left endpoint is excluded and the right is included. A quick test confirms it: 2 fails the inequality, 8 passes, 7 passes.

The third example puts the two together, an open dot at 0 shaded to the right and a separate open dot at 5 shaded to the right, yielding (0, ∞) as a single interval only if there is no gap. If there is a gap at 5 because the second stretch starts at 5, the answer is (0, 5) βˆͺ (5, ∞), which can also be written as all reals except 5. That equivalence is worth checking computationally: substitute x = 5 into the union and confirm it fails both pieces.

Run each conversion twice, once by reading the dots and once by testing a number. Misreading a single dot costs the entire answer, and the test pass catches it in about ten seconds. The bracket, in the end, is just the dot written in ink.

When the Interval Is an Answer to a Real Problem

Domain and range questions from algebra and calculus often arrive already drawn, and the grader wants interval notation rather than inequalities. For a function graphed with a hollow dot at x = βˆ’2 and solid shading to the right, the domain is (βˆ’2, ∞). For a parabola opening upward with a vertex at (1, 3), the range is [3, ∞), because the vertex value is included.

Context changes answer conventions. A word problem about hours studied cannot use negative values, so an algebraic domain of (βˆ’βˆž, ∞) is often restricted to [0, ∞) for the physical situation. State which domain you mean if both appear β€” the mathematical one and the contextual one are different answers to different questions.

Graphing calculators and online homework systems disagree on some edge cases. Reversed brackets, unions typed with a comma instead of the cup symbol, and the word "infinity" typed out are all common causes of entered answers being marked wrong even when the mathematics is right. Check the platform's accepted syntax before the deadline.

Common Core state standards introduce interval notation in high school function work, typically in Algebra I or Algebra II depending on the state's sequencing, so the exact grade varies. If a student is seeing it earlier, it is usually in a pre-algebra enrichment unit rather than a standard requirement.

FAQ

Can a bracket ever touch infinity?

No. Infinity always takes a parenthesis because it is not a number you can reach, and most textbooks treat "∞" as a direction rather than a value. Writing [2, ∞] is a notation error even though the intended meaning is clear. The rule holds for negative infinity as well.

Is 0.999 recurring equal to 1, and does that change anything here?

Yes, 0.999... equals 1, and that identity sometimes surfaces in limit arguments about intervals, but it does not change bracket rules. What changes an answer is whether the endpoint value is included, not how the value is written. Keep the two questions separate.

Why do some textbooks write a backward bracket for an excluded endpoint?

A few texts use ]a, b[ or [a, b) with the bracket facing backward to mean an open interval, following older notation conventions. It is a regional and edition-level variation, not a universal standard. Confirm the convention in the textbook's notation appendix before an exam.

What if the number line has two separate shaded stretches that touch at one point?

If the point where they touch is included by at least one stretch and every value between the outer bounds is shaded, the pieces merge into a single interval. If the touching point is excluded from both, the union keeps two intervals with an open parenthesis at that value on both sides. Test the touching point directly in both pieces.

Does interval notation need to be in order?

Left endpoint first, right endpoint second is the standard, and most graders expect it. A union is commutative, so (βˆ’3, 0] βˆͺ (2, 5] and (2, 5] βˆͺ (βˆ’3, 0] describe the same set, but the ordered version reads more easily. When in doubt, order it.

How many brackets does a union need?

Two intervals need four characters, two for each end: (βˆ’3, 1] βˆͺ [4, ∞). An answer with three brackets or parentheses is missing one, and a grader scanning the ends will catch it. Count the symbols before submitting.

Can an interval have just one endpoint listed?

Yes, for a ray: (3, ∞) lists only the finite end. Some teachers prefer that students write the unbounded side as infinity rather than omitting it, because omitting makes the notation ambiguous with a set listing. Follow the format used in class examples.

What is the difference between interval notation and set-builder notation?

Interval notation compresses the answer into brackets and endpoints, while set-builder notation writes the condition, such as {x | x β‰₯ 3}. They describe the same set and are interchangeable, though interval notation is more compact for continuous ranges. Inequalities are a third way to express the same idea.