What Are One-Sided Limits?
A one-sided limit examines what a function approaches when x closes in on a point from a single direction. The left-hand limit, written limx→c⁻ f(x), uses only x-values smaller than c; the right-hand limit, limx→c⁺ f(x), uses only x-values larger than c. The superscript minus or plus marks the direction of approach — a notation detail that carries the entire meaning.
One-sided limits matter because the ordinary two-sided limit is defined in terms of them: limx→c f(x) exists exactly when both one-sided limits exist and are equal. When the two sides disagree — like |x|/x, which approaches −1 from the left and +1 from the right of zero — the two-sided limit simply does not exist, yet each one-sided limit is perfectly well-defined and informative. Functions defined piecewise, functions with jumps, and functions whose domain ends at a boundary (like √x at 0) all demand one-sided analysis.
How to Use This Calculator
Enter a function and the point c. The calculator samples the function at x-values approaching c from below (c − 0.1, c − 0.01, … down to c − 0.00001) and separately from above, displays both approach tables, and reads off each one-sided limit from where the samples settle. It then compares the two: if they agree on a finite value, it reports the two-sided limit; if they differ or diverge, it explains why the two-sided limit fails to exist. The graph shows the function around c so you can see the behavior each table describes.
Worked Example
Consider f(x) = |x|/x at c = 0. For any negative x, |x| = −x, so f(x) = −x/x = −1; for any positive x, f(x) = x/x = +1. The approach tables therefore lock onto constant values immediately:
limx→0⁻ |x|/x = −1 and limx→0⁺ |x|/x = +1
Since −1 ≠ 1, the two-sided limit at 0 does not exist — the graph makes it obvious, jumping across the gap from −1 to +1 as x crosses zero. This is the canonical jump discontinuity.
Reading One-Sided Results
| Left limit | Right limit | Two-sided conclusion |
|---|---|---|
| L | L (same) | Limit exists and equals L |
| L₁ | L₂ ≠ L₁ | Does not exist (jump) |
| −∞ | +∞ | Does not exist (infinite, opposite signs — e.g. 1/x) |
| +∞ | +∞ | Does not exist as a number, but written lim = +∞ (e.g. 1/x²) |
| exists | undefined (domain ends) | Only the one-sided limit is meaningful (e.g. √x at 0) |
Common Mistakes to Avoid
- Reading the ⁻ and ⁺ superscripts as signs of c. x → 2⁻ means "approaching 2 from below" (through 1.9, 1.99, …), not "approaching −2." The sign marks direction, not the point.
- Concluding "does not exist" too quickly. A function undefined at c can still have both one-sided limits equal — like (x²−1)/(x−1) at x = 1, where both sides approach 2 despite the hole.
- Averaging a jump. If the left limit is −1 and the right is +1, the two-sided limit is not 0 — it does not exist. There's no averaging rule for disagreeing sides.
- Forgetting domain boundaries. For √x at 0, only the right-hand limit makes sense — there are no function values to the left to sample.
Real-World Applications of One-Sided Limits
Real systems are full of quantities that jump: postal rates step up at weight thresholds, tax rates change at bracket boundaries, and utility prices switch between peak and off-peak at fixed clock times. Near any such threshold, the question "what does it cost just below versus just above?" is a pair of one-sided limits, and the difference between them is the size of the jump — exactly what a pricing analyst or an affected customer needs to know. Piecewise-defined engineering controls (a thermostat's on/off switching, gear-shift points in a transmission) are analyzed the same way.
In physics and signal processing, one-sided limits define behavior at switching instants: the voltage just before and just after a circuit closes, or a signal's value approaching a discontinuity, are written f(t⁻) and f(t⁺) and appear throughout Laplace-transform analysis of circuits and control systems. And within mathematics itself, one-sided limits define one-sided derivatives, which detect corners: |x| has a left-derivative of −1 and a right-derivative of +1 at zero, and their disagreement is precisely why no tangent line exists there.
Frequently Asked Questions
Can a one-sided limit exist when the two-sided limit doesn't?
Yes — that's the typical situation at a jump. Each side can settle on its own value; the two-sided limit fails only because the two values differ. One-sided limits are strictly more forgiving than two-sided ones.
What notation distinguishes left from right limits?
A superscript on the approach value: x → c⁻ for the left (values below c) and x → c⁺ for the right (values above c). Some textbooks instead write limx↑c and limx↓c, with the arrows indicating approach from below or above.
How do one-sided limits relate to continuity?
Continuity at c requires the two-sided limit to exist and equal f(c) — which means both one-sided limits must exist, agree with each other, and agree with the function's value. The Continuity Calculator runs that full three-part test and classifies any failure.
What is "continuity from the right"?
A function is right-continuous at c if limx→c⁺ f(x) = f(c), using only the right-hand limit. It's the natural notion at domain boundaries — √x is right-continuous at 0 — and it's the standard convention for step functions in probability (cumulative distribution functions are defined to be right-continuous).
Why does the calculator sample at shrinking distances instead of solving symbolically?
Sampling from each side mirrors the definition of a one-sided limit directly and works for any expression, including ones with jumps and asymptotes that resist symbolic treatment. The shrinking-step tables also show the settling behavior rather than just asserting it — for a symbolic two-sided approach with L'Hôpital's rule, use the main Limit Calculator.
Do one-sided limits obey the usual limit laws?
Yes — sums, products, and quotients of one-sided limits follow the same rules as two-sided ones, provided each piece's one-sided limit exists and denominators stay nonzero. That means you can compute a complicated left-hand limit by breaking the expression apart, resolving each part from the left, and recombining — the same algebra of limits you already know, just with a direction attached throughout.