Differentials are a compact way to describe how a small change in an input produces an approximate change in an output. In calculus, dx represents a small chosen change in x, while dy represents the change predicted by the tangent line to y = f(x). This idea matters because many real measurements and calculations involve tiny changes, not perfectly exact values.
Differentials turn the derivative into a practical tool for estimation.
At a point x = a, the derivative f'(a) gives the slope of the tangent line, so the differential satisfies dy = f'(a) dx. The actual change in the function is Δy = f(a + dx) - f(a), which is usually close to dy when dx is small and the curve is smooth. This leads to the linear approximation f(a + dx) ≈ f(a) + f'(a) dx.
Differentials are also used to estimate propagated error, such as how uncertainty in a radius affects the uncertainty in area or volume.
Understanding Calculus: Differentials
A differential is best understood as part of a local straight-line model. Near one input value, a smooth curve can be replaced briefly by its tangent line. The tangent line is not the original curve.
It is a simpler model that agrees with the curve at one point and has the same immediate steepness there. Its accuracy depends on curvature. If the graph bends strongly, the tangent line separates from the graph more quickly as the input moves away from the chosen point.
This is why a small input change is not enough by itself. The function must behave smoothly in that nearby region, and the change must be small compared with the scale over which the graph bends.
Units give differentials an important physical meaning. If distance is measured in metres and time in seconds, then a derivative of distance with respect to time has units of metres per second. Multiplying that rate by a small time change gives a predicted distance change in metres.
The units work like a built-in check on the calculation. For a circle, a small increase in radius produces an area increase of about two times pi times the radius times the radius change. If the radius is ten centimetres and its change is one tenth of a centimetre, the estimated area change is about two times pi square centimetres.
The exact area change is slightly larger because the radius change is squared in the full expansion. That extra part shows the effect of curvature.
Differentials are especially useful when measurements are uncertain. A ruler, sensor, or digital scale cannot usually give a perfectly exact reading. The input uncertainty can be treated as a small positive or negative change.
The derivative then tells how sensitive the output is to that uncertainty. A steep derivative means a small input error can create a larger output error. A derivative near zero means the output is less sensitive at that location.
This matters in engineering, laboratory work, map measurements, and manufacturing. It is important to remember that an estimated error is often a size, not a direction. Taking the absolute value gives the largest likely size predicted by the linear model.
When learning this topic, keep the chosen point visible throughout the work. First find the derivative. Next evaluate it at the original input value.
Then multiply by the stated small change, keeping the sign and units. Do not substitute the new input into the derivative unless the task specifically asks for a new local estimate. Compare the estimate with an exact calculation when one is easy to make.
The difference reveals how well the tangent-line model worked. Watch out near corners, cusps, vertical tangents, and discontinuities.
At such points, an ordinary derivative may not exist, so the usual differential method cannot be trusted. A graph can help because it makes the gap between the curve and its tangent line easier to see.
Key Facts
- dx is a small change in the input x, often chosen by the user or measurement situation.
- dy is the tangent-line change in y, given by dy = f'(x) dx.
- The actual change is Δy = f(x + dx) - f(x), which is not always equal to dy.
- Linear approximation: f(a + Δx) ≈ f(a) + f'(a) Δx.
- For small changes, Δy ≈ dy when f is differentiable and Δx is close to 0.
- Estimated error in a function value can be written as |dy| = |f'(x)| |dx|.
Vocabulary
- Differential
- A differential is an expression that estimates a small change in a quantity using a derivative.
- dx
- dx is a small change in the independent variable x.
- dy
- dy is the approximate change in y predicted by the tangent line, calculated as dy = f'(x) dx.
- Tangent line
- A tangent line is the line that touches a curve at a point and has slope equal to the derivative there.
- Linear approximation
- Linear approximation estimates a function near a point by using the value and slope of the tangent line.
Common Mistakes to Avoid
- Treating dy and Δy as always equal is wrong because dy is the tangent-line estimate while Δy is the actual change on the curve.
- Forgetting to evaluate the derivative at the starting point is wrong because dy = f'(a) dx uses the slope at x = a, not at a random nearby value.
- Using a large dx without checking accuracy is wrong because linear approximation becomes less reliable as the input change gets farther from the base point.
- Ignoring units in differentials is wrong because dx and dy carry the units of their variables, and the derivative carries output units per input unit.
Practice Questions
- 1 For f(x) = x^2 at x = 3, use differentials to estimate the change in y when dx = 0.1. Then compare with the actual change Δy.
- 2 The radius of a circle is measured as r = 10 cm with a possible error of dr = 0.05 cm. Use dA = 2πr dr to estimate the possible error in the area.
- 3 A curve is concave up near x = a and dx is positive. Explain whether the tangent-line differential dy is likely to overestimate or underestimate the actual change Δy.