Did you know? A graph can look perfectly ordinary until we turn every non-zero height into its reciprocal. A point at height two moves to height one-half. A point at height one-half moves to height two. And a point at height zero cannot be carried across at all, because 1/0 is undefined. This last detail is why the graph of y = 1/f(x) often shoots up or down near a zero of f.
For students searching for a JC1 H2 Mathematics Tutor Bukit Timah, reciprocal graph transformations are a useful early test of whether functions, domains and graphs have become connected ideas. The Singapore–Cambridge H2 Mathematics 9758 syllabus for 2027 explicitly includes relating y = 1/f(x) to y = f(x). That does not mean every school teaches the chapter at the same time, but the mathematical reasoning belongs to the wider A-Level learning route.
This guide is deliberately about one question: how can a learner sketch or reason about the reciprocal graph without blindly replacing a curve with a memorised hyperbola? We work from the point mapping, then treat zeros, asymptotes, signs, missing domains, a few awkward examples, calculator checks and independent problems. We do not repeat the full inverse-function chapter already covered in the eduKateSG JC1 functions guide.
Quick answer for a reader with thirty seconds
The reciprocal graph takes every allowed point (x, y) on the original curve, provided y is not zero, and changes it to (x, 1/y). The x-coordinate does not move. The new height depends entirely on how far the original height lies from zero and on its sign.
When f(x) is positive, 1/f(x) is positive. When f(x) is negative, 1/f(x) is negative. A high positive value such as four becomes a small positive value one-quarter. A small positive value such as one-quarter becomes four. Negative numbers keep their negative sign under reciprocation.
If a continuous original graph approaches zero at an input a while staying non-zero nearby, then the magnitude of the reciprocal tends to become very large. This commonly produces a vertical asymptote x = a for the reciprocal graph. We must still check the exact original domain and the behaviour on each side, rather than declaring every visually missing point an asymptote.
- Find the original domain of f before doing any reciprocal transformation.
- Exclude every x for which f(x) = 0, because the reciprocal is undefined there.
- Copy the original x-coordinate, but replace every non-zero y-value with 1/y.
- Keep track of the sign of f on each interval separated by zeros or exclusions.
- Examine what f does near zero, near its original undefined inputs and as x becomes large in magnitude.
- Sketch using transformed landmarks, interval signs and asymptotic behaviour instead of guessing from a generic shape.
First principles: the reciprocal is not an inverse function
Two symbols repeatedly collide in a student’s notes. The expression 1/f(x) is a reciprocal of an output. The expression f⁻¹(x) is an inverse function, when such a function exists. They answer different mathematical questions.
A reciprocal asks, “What number multiplied by the original output gives one?” If f(3) = 4, then 1/f(3) = 1/4. We still put three into the new function. The input has not been reversed or recovered.
An inverse function instead asks, “Which original input produced this output?” If f(x) = 2x + 1, the inverse, on a suitable domain, is f⁻¹(x) = (x − 1)/2. That is not the same as 1/(2x + 1). The first reverses a mapping; the second changes the height of the output.
Try a deliberately simple check. For f(x) = 2x + 1, f(1) = 3. The reciprocal at input one is 1/3. The inverse at input one is zero, because zero is the original input that produces output one. Both calculations are correct, but they solve different problems.
In a tutorial, I would have the learner say these meanings aloud before moving to a graph. If the meaning is unstable, memorising that an inverse reflects across y = x and a reciprocal does something else will not transfer reliably to an unfamiliar curve.
A point-mapping table you can build yourself
Suppose a set of labelled points on y = f(x) has heights −4, −1, −1/2, 1/2, 1 and 4. Their reciprocal heights are −1/4, −1, −2, 2, 1 and 1/4. This tiny table contains nearly everything we need to understand the graph’s vertical changes.
The point (2, 4) becomes (2, 1/4). The point (−1, −1/2) becomes (−1, −2). The point (0, 1) stays at (0, 1). No reflection across a diagonal line is involved. Every transformation happens vertically at the same x-coordinate.
Values with magnitude bigger than one move towards the x-axis after reciprocation. Values with magnitude smaller than one move away from it. The heights 1 and −1 are fixed by the transformation because each is its own reciprocal.
Notice that the original curve might not contain every possible height in this list. We are not inventing points. We transform only the actual points of the original curve, subject to its domain. If the original domain omits x = 5, no reciprocal manipulation may create a new point at x = 5.
An important difference between distance and sign
There is a special subtlety with negative values. A learner might see y = −1/4 and think “the point moves up because one-quarter becomes four.” In fact the reciprocal is −4: it moves farther down, away from the x-axis. The correct statement is that the magnitude increases when the absolute value of the original height lies between zero and one.
It is safer to separate two questions. First, is the original output positive or negative? Second, is its absolute value greater or less than one? The sign determines the half-plane; the magnitude determines whether the point moves nearer to or farther from the horizontal axis.
One sentence makes the idea memorable: reciprocation preserves the sign and reverses the relationship between large and small non-zero magnitudes. This is more precise than saying “the graph flips.”
Worked example one: from a line to a hyperbola
Let f(x) = x − 2 for all real x. Its graph is a straight line with an x-intercept at (2, 0). The reciprocal is g(x) = 1/(x − 2). This new function is defined only when x ≠ 2.
At x = 3, the original output is one, so g(3) = 1. At x = 4, the original output is two, so g(4) = 1/2. At x = 1, the original output is minus one, so g(1) = −1. At x = 0, the original output is minus two, so g(0) = −1/2.
What happens close to x = 2? For x = 2.1, the denominator is 0.1 and g(x) = 10. For x = 1.9, the denominator is −0.1 and g(x) = −10. As we approach two from the right, the reciprocal grows without bound positively. From the left it decreases without bound negatively.
Therefore x = 2 is a vertical asymptote. The horizontal asymptote is y = 0 because the denominator grows in magnitude as x moves far to either side and the reciprocal approaches zero.
The two branches are separated by the excluded input x = 2. The right branch is positive, and the left branch is negative. A good sketch marks both asymptotes and locates at least one point in each branch before drawing a smooth curve.
How the line’s zero becomes an asymptote
The original line meets the x-axis at x = 2. A reciprocal output would need to evaluate 1/0 there, which is impossible. Nearby, a tiny denominator produces a very large magnitude. That is why the new graph cannot simply pass through or smoothly touch the old zero.
But the reasoning is about the behaviour as x approaches the value. An undefined input does not, by itself, prove that a graph has a vertical asymptote. For example, a function with a removable hole and finite nearby values may merely remain missing at one point after a certain transformation. We will examine that distinction below.
The careful examination statement is: “As x approaches two from each allowed side, 1/(x − 2) becomes unbounded, so x = 2 is a vertical asymptote.” That gives the mathematical reason rather than merely repeating a rule.
Worked example two: the reciprocal of a quadratic
Take f(x) = x² − 4. The original parabola opens upward, has zeros at x = −2 and x = 2, has its minimum (0, −4), and is symmetric about the y-axis. Its reciprocal is g(x) = 1/(x² − 4).
The domain of g is all real x except −2 and 2. On x < −2 and x > 2, x² − 4 is positive, so the reciprocal is positive. Between −2 and 2, the denominator is negative and the reciprocal is negative.
At x = 0, g(0) = −1/4. At x = 3, g(3) = 1/5. At x = −3, g(−3) = 1/5. The equal values on opposite sides show that the reciprocal retains even symmetry in this example.
Close to x = 2 from the left, x² − 4 is a small negative number, so g dives downward. From the right, the denominator is a small positive number, so g shoots upward. The same reasoning around x = −2 gives positive infinity from the left and negative infinity from the right.
Thus the sketch has two vertical asymptotes, x = −2 and x = 2, and horizontal asymptote y = 0. The central branch is negative and symmetric, with point (0, −1/4). The two outer branches are positive.
Why the central branch must remain below the axis
It is tempting to copy the top half of a standard hyperbola sketch and place it between the asymptotes. The sign chart immediately corrects this error: when −2 < x < 2, the expression x² − 4 is negative, so its reciprocal must be negative. No correctly drawn point in that interval can have a positive y-coordinate.
This is why the sign table comes before the polished picture. It is a structural constraint. A curve that breaks it cannot be saved by making the asymptote labels look neat.
A learner can independently verify the sign with the easy point x = 0. The reciprocal there is −1/4. That tells us which side of the x-axis contains the entire central branch, because the denominator never crosses zero within the interval.
Worked example three: a repeated root behaves differently
Let f(x) = (x − 1)². The original quadratic touches the x-axis at x = 1 without crossing it. Its reciprocal is g(x) = 1/(x − 1)², defined for all x except one.
Because the denominator is a square, it is positive for every allowed input. The reciprocal is therefore always positive. As x approaches one from either side, the denominator approaches zero through positive values and g approaches positive infinity on both sides.
The vertical asymptote is x = 1, but the branches now rise on both sides instead of one rising and the other falling. That is different from the earlier linear denominator, whose sign changed at its root.
At x = 0 or x = 2, g = 1. At x = −1 or x = 3, g = 1/4. These paired points show symmetry about the vertical line x = 1, not necessarily the y-axis.
The deeper principle is that the multiplicity and sign behaviour of the original zero influence which direction the reciprocal branch takes. A simple sign-changing zero commonly produces opposite infinity signs; a squared, non-sign-changing zero can produce the same infinity sign on both sides.
Worked example four: a function with no real zeros
Not every reciprocal graph needs a vertical asymptote. Consider f(x) = x² + 1. The original curve is always at least one unit above the x-axis. It never has a real zero, so 1/f(x) = 1/(x² + 1) is defined for every real x.
The reciprocal is positive everywhere. At x = 0, its value is one. At x = ±1, its value is one-half. At x = ±3, its value is one-tenth. As |x| grows, it approaches zero from above.
The graph is an even bell-like curve with maximum at (0, 1), not a function that shoots up around some missing input. Its horizontal asymptote is y = 0, and it has no vertical asymptotes.
This is a useful counterexample to the over-simple slogan “reciprocal graphs always have vertical asymptotes.” The more accurate statement is that the reciprocal can become unbounded near points where its denominator approaches zero, provided those points can be approached through the original domain.
Range is a second question, not an afterthought
The original f(x) = x² + 1 has range [1, ∞). Reciprocation turns that range into (0, 1], because the reciprocal of one is one and the reciprocals of larger positive numbers fall strictly between zero and one.
The new curve can get arbitrarily close to zero, but it never equals zero. The output 1/(x² + 1) cannot be zero, because its numerator is the constant one. This distinction between an approached value and an attained value matters when stating the range.
A student who writes 0 ≤ y ≤ 1 has included an output the function cannot achieve. A properly stated answer is 0 < y ≤ 1. Thinking about the original range supplies an efficient check.
Zeros versus undefined points versus asymptotes
Here is the distinction that makes this topic more than a drawing trick. A zero of f is a defined input where f(x) = 0. An undefined point is an input not in f’s domain. A vertical asymptote concerns the values a function takes as allowed inputs approach a particular x-value.
These concepts may coincide in a reciprocal graph but need not. If a continuous f has a simple real zero, the reciprocal commonly develops a vertical asymptote. If f is undefined because an algebraic formula has a cancelled factor, the reciprocal may remain undefined there without becoming unbounded.
Consider f(x) = (x² − 1)/(x − 1), but keep the original domain restriction x ≠ 1. On that domain, f(x) = x + 1. The reciprocal is 1/f(x) = 1/(x + 1), with the original exclusion x ≠ 1 still in force.
Therefore its domain excludes both x = −1, where f is zero, and x = 1, where f was never defined. At x = −1 the reciprocal is unbounded nearby, so there is a vertical asymptote. At x = 1 the nearby reciprocal values approach 1/2, so there is a hole at (1, 1/2), not a vertical asymptote.
A student who simplifies the original expression and forgets x ≠ 1 may add an invented point. A student who calls every missing x an asymptote may mistake a finite hole for infinite behaviour. The exact repair is to preserve the original domain and inspect nearby values separately.
What happens to horizontal asymptotes?
If f(x) approaches a non-zero constant c as x approaches positive or negative infinity, then 1/f(x) approaches 1/c, provided f remains non-zero eventually. Thus a non-zero horizontal asymptote y = c of f can correspond to y = 1/c for its reciprocal.
For example f(x) = 2 + 1/x on domain x ≠ 0. As x becomes very large in magnitude, f approaches two. The reciprocal 1/f(x) = x/(2x + 1), on the original permitted domain where also 2x + 1 ≠ 0, approaches one-half.
The original function has a horizontal asymptote y = 2. Its reciprocal has horizontal asymptote y = 1/2. That makes sense: far away, the original outputs are near two, so their reciprocals are near one-half.
If f(x) instead grows without bound in magnitude, then 1/f(x) may approach zero. For x² + 1 we saw exactly that. If f(x) approaches zero, the reciprocal may instead become unbounded; the behaviour must be examined with attention to sign.
Can y = 1/f(x) ever cross the x-axis?
No, not where it is defined. The equation 1/f(x) = 0 has no solution for any finite non-zero real denominator. This is one of the most reliable checks in the chapter.
A reciprocal graph may approach the x-axis closely, as with 1/(x² + 1), but it cannot meet it. If a sketch shows an ordinary x-intercept for a reciprocal with numerator one, there is a problem.
What about a point where the original f(x) tends towards infinity? Its reciprocal may tend towards zero, but that gives an asymptotic or limiting behaviour—not a guarantee that a point with y = 0 belongs to the new graph. Check the exact domain.
This conclusion also helps with solving equations. If the question asks for intersections of y = 1/f(x) with the line y = 0, the answer is none. If the question asks for intersections with y = 1, those occur exactly where f(x) = 1 and the original function is defined.
Why a sign chart is better than an artistic guess
For g(x) = 1/(x² − 4), draw a line marking the critical x-values −2 and 2. Check the sign of the denominator on each interval. The sign of the reciprocal is identical to the sign of the denominator because the numerator is one.
On the left outer interval the graph is positive; in the middle it is negative; on the right outer interval it is positive. Then check the limiting behaviour close to each boundary. These two pieces—interval sign and boundary behaviour—determine much of the correct qualitative shape.
This method scales to f(x) with repeated roots or with known signs supplied by a sketch. It is also useful where explicit algebra is difficult or where only a graph is given.
The goal is not to draw dozens of test points. It is to choose a small set of decisive facts: exclusion locations, sign intervals, transformed landmarks, turning behaviour where justified, and limits near significant boundaries. A graphing calculator can then serve as a check rather than the source of an unexplained picture.
A parent-friendly diagnostic conversation
Suppose a student has written “the graph has two vertical asymptotes because it is a reciprocal.” A tutor should ask for the original f(x), not accept the sentence as evidence. Which values make f(x) zero? Are they in the domain? Does the reciprocal blow up near them? That turns a slogan into a mathematical argument.
Next ask the student to explain why |f(x)| and 1/f(x) are different. The first takes absolute values of outputs, making the graph non-negative. The second takes reciprocal outputs, preserving positive or negative signs and excluding zeros.
Then present f(x) = x² + 1 and ask whether a vertical asymptote exists. A student who confidently says “no, because the denominator never vanishes and stays at least one” has understood more than a student who can only copy a standard hyperbola.
For a family in Bukit Timah, this sort of five-minute conversation can reveal whether the next lesson should address function notation, domain restrictions, rational-expression signs or graph interpretation. The correct repair depends on that diagnosis.
A three-stage tutor progression
Stage one: transform a list of points
Begin with an original graph or a table containing easy output heights: ±1/2, ±1, ±2 and ±4. Ask for reciprocal outputs while keeping each x-coordinate unchanged. Include zero as a deliberate forbidden case. Make the student explain what happens to signs.
At this stage, do not introduce multiple asymptotes or complicated rational functions. The target is understanding the mapping, and the check is whether the student can perform it consistently on negative and positive values.
Stage two: connect roots and asymptotes
Use f(x) = x − 2, then f(x) = (x − 2)². Both have a zero at two, but their signs behave differently across the zero. Ask the student to choose points immediately to the left and right and predict whether the reciprocal becomes large positive or large negative.
Only after that prediction do we sketch. The learner should be able to justify the branches using a denominator sign, not simply draw two curves they remember from a formula sheet.
Stage three: preserve domain and transfer to an unseen graph
Introduce the removable-hole example, then a graph with no zeros, then an unseen quadratic with two roots. Ask for a written domain, sign intervals, transformed landmarks and a sketch. Finally, compare the student’s explanation with the calculator picture.
The learning objective is not “correctly draw five reciprocal curves today.” It is “have a reliable method for a reciprocal graph the learner has never seen.” That is the standard of independent mathematical understanding.
Worked example five: reciprocal of a shifted positive quadratic
Consider f(x) = (x − 3)² + 4. Its minimum is four at x = 3, and it is positive for every real x. Its reciprocal is g(x) = 1/[(x − 3)² + 4].
The reciprocal has maximum 1/4 when the denominator is smallest. Thus the maximum point is (3, 1/4). The function is positive everywhere, symmetric about x = 3, and approaches zero as x moves far from three. Its range is (0, 1/4].
There is no vertical asymptote because the denominator never approaches zero; it is at least four. The horizontal asymptote is y = 0.
This example connects the reciprocal transformation with a useful principle about extrema: when positive outputs are reciprocated, a smaller positive denominator gives a larger reciprocal. It is not a universal rule for negative functions or for functions that cross zero without additional reasoning. The positivity condition matters.
Worked example six: a reciprocal with an existing hole
Let f(x) = (x − 2)(x + 3)/(x − 2), but define it from its original algebraic expression on domain x ≠ 2. Then f(x) = x + 3 for allowed x.
Its reciprocal g(x) = 1/(x + 3) is valid only when x ≠ 2 and x ≠ −3. The first exclusion is inherited from f; the second comes from f(x) = 0.
At x = 2, g has a missing value that would otherwise be 1/5. Near x = 2 the reciprocal remains finite. So this is a hole, not an asymptote. At x = −3, the denominator approaches zero and g becomes unbounded, creating a vertical asymptote.
The graph must therefore show a rational hyperbola with vertical asymptote x = −3, horizontal asymptote y = 0, and an additional hole at (2, 1/5).
This is not merely algebraic pedantry. It is precisely the kind of changed-condition case that reveals whether the student has learned an actual function definition or just symbol cancellation.
Twelve practice tasks with short, checked answers
Question 1. A point (4, 5) lies on f. Where does it go on 1/f? Answer: (4, 1/5), because its x-coordinate stays four and the height is reciprocated.
Question 2. A point (−2, −1/4) lies on f. What is its reciprocal point? Answer: (−2, −4).
Question 3. Can a reciprocal graph contain (3, 0)? Answer: No. The equation 1/f(3) = 0 cannot hold for a finite non-zero value of f(3).
Question 4. For f(x) = x + 4, where is the reciprocal undefined? Answer: x = −4, where the original output is zero.
Question 5. On which side of x = −4 is 1/(x + 4) positive? Answer: On x > −4.
Question 6. For f(x) = x² + 9, identify vertical asymptotes of 1/f. Answer: None, because f(x) is at least nine for all real x.
Question 7. What is the maximum value of 1/(x² + 9)? Answer: 1/9 at x = 0.
Question 8. For f(x) = x² − 9, state the forbidden x-values of 1/f. Answer: x = −3 and x = 3.
Question 9. For f(x) = −(x − 1)², what sign does its reciprocal have wherever defined? Answer: Negative, because the denominator is strictly negative except at x = 1 where it is zero and excluded.
Question 10. Is f⁻¹(x) the same as 1/f(x)? Answer: No. f⁻¹ reverses a one-to-one mapping; 1/f reciprocates a non-zero output.
Question 11. An original function tends to two as x tends to infinity and remains non-zero eventually. What happens to its reciprocal? Answer: It tends to one-half.
Question 12. An original rational function has a removable hole where its nearby output tends to five. If the reciprocal exists on nearby points, what is its nearby output limit? Answer: One-fifth, although the hole remains excluded from the domain.
A useful independent check is to have the learner choose two of the questions and give their reason rather than only the numerical answer. The explanations about domain and near-zero behaviour are the ones that most strongly distinguish understanding from memorisation.
Longer transfer challenge: infer a graph from only its signs
Suppose an unknown continuous function f is defined on all real numbers, has exactly two simple zeros at −1 and 3, is positive for x < −1 and x > 3, and negative for −1 < x < 3. You are not given its exact formula.
What can you assert about g(x) = 1/f(x)? First, its domain excludes −1 and 3. Second, g has the same sign as f on every permitted interval. Third, near each simple sign-changing zero, the reciprocal becomes large in magnitude, with the sign determined by the side from which we approach.
But you cannot determine the exact height of a point without knowing the value of f there. You also cannot deduce a particular horizontal asymptote as x tends to infinity unless you know how f behaves there.
This is an important lesson in evidence: a graph sketch can be partly determined by qualitative information without being uniquely determined. An H2 Maths student should state what follows and what remains unknown rather than inventing extra structure.
If one additional fact says f(x) → +∞ as x → ±∞, then g(x) → 0 from above on the outer tails. That new information is enough to justify the horizontal asymptote y = 0 there. The change from uncertain to determined is visible.
When a graphing calculator helps—and when it can mislead
An approved graphing calculator can reveal whether branches rise or fall, and it can help check a formula such as 1/(x² − 4). Yet a plot may connect points near a vertical asymptote with an apparent line, especially if the viewing window or plotting algorithm samples too coarsely.
A student should therefore mark the forbidden x-values before tracing the calculator graph. If the screen shows an implausible line passing across x = 2, do not override the algebraic fact that the denominator is zero there.
Likewise, an extremely small reciprocal value might round to zero on the display. That does not produce a genuine x-intercept. The exact numerator remains one, so the mathematical output is never zero at any allowed finite input.
The 2027 official H2 Mathematics syllabus discusses graphing calculator use and expects mathematical steps where the question calls for them. The calculator is a valuable diagnostic companion, not a license to ignore domain statements or exact reasoning.
Frequently asked questions
Do zeros of f always become vertical asymptotes of 1/f?
For a continuous f that approaches zero through non-zero nearby values, the reciprocal commonly becomes unbounded and produces a vertical asymptote. A full answer checks the original domain and the limiting behaviour rather than treating the statement as an unqualified rule for arbitrary discontinuous functions.
Is a hole the same as an asymptote?
No. A hole is a missing point; nearby values may tend to a finite number. A vertical asymptote is identified by unbounded behaviour as the input approaches a value. The reciprocal of a rational function can show both types of exclusion.
Can 1/f(x) be negative?
Yes, whenever the original f(x) is negative and non-zero. Reciprocation preserves the sign of each non-zero real output.
Can 1/f(x) have an x-intercept?
No. Its numerator is one. There is no allowed finite x for which 1/f(x) equals zero.
What happens to a point where f(x) = 1 or f(x) = −1?
The point keeps its output because 1/1 = 1 and 1/(−1) = −1. These are useful fixed-height landmarks.
Is the graph of 1/f(x) a reflection of the graph of f(x)?
Not generally. It maps (x, y) to (x, 1/y) for non-zero y. That is not an ordinary reflection across the x-axis, y-axis or y = x.
Why is this useful beyond one JC chapter?
The same habits—checking domains, signs, limiting values and transformed coordinates—support rational function problems, inequalities, calculus and later mathematical modelling. The point is transferable reasoning rather than merely one graph family.
Your next step and related reading
If reciprocal notation remains uncertain, revise the JC1 H2 functions, domains and inverses guide. If the difficulty is drawing transformed curves, visit why f(x + a) moves left and why f(|x|) differs from |f(x)|. For another precise graph interpretation, see the guide to rational inequality sign charts.
For subject scope, use the Singapore–Cambridge 2027 H2 Mathematics 9758 syllabus rather than assuming an individual school’s JC1 chapter order. Families seeking current local arrangements can visit Bukit Timah JC1 tuition.
The immutable eduKateSG Secondary 1 Mathematics small-group teaching reference illustrates a diagnosis → explanation → practise → independent-check rhythm. It is a different level and its class details should not be treated as promises about a JC1 programme.
The last idea is the one to carry into the next unseen question: a reciprocal graph inherits the original input rules, transforms each non-zero height, and makes small allowed denominators matter enormously. That thought is more useful than any memorised sketch.