Limits, Continuity & Differentiability
Learning Objectives
- Understand the fundamental concepts of Limits, Continuity & Differentiability
- Apply key formulas and techniques to solve problems
- Practice with exam-level questions to build speed and accuracy
1. Limits and Standard Results
1.1 Definition of a Limit
The concept of a limit captures the behaviour of a function as its argument approaches a particular value, without necessarily requiring the function to be defined at that value. Formally, we say that the limit of f(x) as x tends to a is L, written limx→a f(x) = L, if for every positive number ε there exists a positive number δ such that whenever 0 < |x − a| < δ the inequality |f(x) − L| < ε holds. This ε‑δ definition makes precise the intuitive idea that f(x) can be made arbitrarily close to L by taking x sufficiently close to a, and it is the foundation upon which all subsequent limit theorems are built.
Because the definition only concerns values of x different from a, the existence of a limit does not depend on the value f(a) itself. Consequently a function may have a limit at a point where it is not defined, or where it is defined but takes a different value. This distinction is crucial when we later discuss continuity, where the limit must equal the function value.
1.2 Algebra of Limits
If the limits of two functions exist individually, then the limit of their sum, difference, product, and quotient (provided the denominator limit is non‑zero) can be obtained by performing the corresponding algebraic operation on the individual limits. In symbols, if limx→a f(x) = L and limx→a g(x) = M, then limx→a [f(x) ± g(x)] = L ± M, limx→a [f(x)·g(x)] = L·M, and limx→a [f(x)/g(x)] = L/M for M ≠ 0. These rules follow directly from the ε‑δ definition and allow us to break down complicated expressions into simpler pieces whose limits are known or easier to evaluate.
The algebra of limits also extends to constant multiples and powers: limx→a [c·f(x)] = c·L for any constant c, and limx→a [f(x)]n = Ln for any positive integer n. When combined with the standard limits discussed next, these rules give a powerful toolkit for evaluating a vast majority of elementary limits encountered in the JEE syllabus.
1.3 Standard Limits and Their Derivations
Several limits appear so frequently that they are treated as standard results. The most important among them is limx→0 (sin x)/x = 1, which can be proved geometrically by comparing the areas of a sector, a triangle, and a larger triangle in the unit circle, or analytically using the squeeze theorem. From this fundamental limit we obtain limx→0 (tan x)/x = 1 and limx→0 (1 − cos x)/x2 = 1/2 by algebraic manipulation and the identity 1 − cos x = 2 sin2(x/2). Another cornerstone is limx→0 (ex − 1)/x = 1, which follows from the definition of the exponential function as its own derivative or from the series expansion ex = 1 + x + x2/2! + … . Similarly, limx→0 (ax − 1)/x = ln a for a > 0, and limx→0 (ln(1+x))/x = 1. These standard limits are the building blocks for evaluating more complex indeterminate forms without resorting to L’Hôpital’s rule at every step.
Knowing the derivations helps a student recognise when a given limit can be transformed into one of these standard forms. For instance, the limit limx→0 (sin 5x)/(3x) is immediately seen as (5/3)·limx→0 (sin 5x)/(5x) = 5/3, because the argument of sine and the denominator can be matched by a simple substitution. Such manipulations are expected to be performed swiftly in an examination setting.
1.4 Limits at Infinity and Infinite Limits
When the variable grows without bound, we speak of limits at infinity. The notation limx→∞ f(x) = L means that for every ε > 0 there exists a number M such that |f(x) − L| < ε whenever x > M. This definition mirrors the finite‑point definition and allows us to discuss horizontal asymptotes of graphs. For rational functions, the limit at infinity is determined by the leading terms of numerator and denominator: if the degree of the numerator is less than that of the denominator the limit is 0; if the degrees are equal the limit is the ratio of leading coefficients; if the numerator’s degree exceeds the denominator’s the limit is infinite (or does not exist in the finite sense).
Infinite limits, written limx→a f(x) = ∞, describe the situation where f(x) grows arbitrarily large in magnitude as x approaches a. They are not limits in the strict ε‑δ sense because ∞ is not a real number, but they are a convenient shorthand for describing vertical asymptotes. The algebraic rules for limits extend to infinite limits with the usual caveats: ∞ + ∞ = ∞, ∞·∞ = ∞, but expressions such as ∞ − ∞ or 0·∞ are indeterminate and require further analysis.
| Property | Limit Law | Continuity Requirement |
|---|---|---|
| Sum | lim (f+g) = lim f + lim g | f+g continuous if f and g continuous |
| Product | lim (f·g) = (lim f)(lim g) | f·g continuous if f and g continuous |
| Quotient | lim (f/g) = (lim f)/(lim g), lim g ≠ 0 | f/g continuous if f,g continuous and g ≠ 0 |
| Composition | lim f(g(x)) = f(lim g(x)) if f continuous at lim g | f∘g continuous if g continuous at a and f continuous at g(a) |
2. Continuity
2.1 Continuity at a Point
A function f is said to be continuous at a point a if three conditions are simultaneously satisfied: f(a) is defined, the limit limx→a f(x) exists, and the limit equals the function value, i.e. limx→a f(x) = f(a). This definition encapsulates the idea that the graph of the function has no break, jump, or hole at a. If any one of the three conditions fails, the function is discontinuous at that point. The ε‑δ formulation of continuity is a direct adaptation of the limit definition: for every ε > 0 there exists a δ > 0 such that |x − a| < δ implies |f(x) − f(a)| < ε. Notice that the condition 0 < |x − a| is dropped because we now require the inequality to hold also at x = a, which is automatically true when the limit equals the function value.
Continuity at a point is a local property; it depends only on the behaviour of the function in an arbitrarily small neighbourhood of that point. This locality is why we can often test continuity by examining the left‑hand and right‑hand limits separately, especially for piecewise‑defined functions.
2.2 Continuity on an Interval
A function is continuous on an open interval (a, b) if it is continuous at every point of that interval. For a closed interval [a, b] we additionally require right‑continuity at a (limx→a⁺ f(x) = f(a)) and left‑continuity at b (limx→b⁻ f(x) = f(b)). The distinction between open and closed intervals matters when applying theorems such as the Intermediate Value Theorem, which guarantees that a continuous function on a closed interval attains every value between its endpoints. The theorem fails on open intervals because the function may approach a value without ever reaching it, as seen with f(x) = 1/x on (0, 1).
Continuity on an interval also implies the function is bounded and attains its bounds (Extreme Value Theorem) provided the interval is closed and bounded. These global consequences are direct results of the local ε‑δ definition combined with the completeness of the real numbers.
2.3 Classification of Discontinuities
When a function fails to be continuous at a point, the discontinuity can be classified into three main types. A removable discontinuity occurs when the limit exists but either the function is not defined at the point or its value differs from the limit; the graph has a hole that can be “filled” by redefining the function value. A jump discontinuity arises when the left‑hand and right‑hand limits both exist but are unequal; the graph makes a finite jump. An infinite (or essential) discontinuity occurs when at least one of the one‑sided limits is infinite or does not exist in any finite sense; the graph typically has a vertical asymptote. Recognising the type of discontinuity is important for determining whether a function can be made continuous by a simple redefinition or whether the discontinuity is intrinsic to the function’s nature.
2.4 Properties of Continuous Functions
Continuous functions enjoy several algebraic properties that mirror the limit laws: the sum, difference, product, and quotient (where the denominator is non‑zero) of continuous functions are continuous. Moreover, the composition of continuous functions is continuous. These properties allow us to build complicated continuous functions from elementary ones such as polynomials, trigonometric functions, exponentials, and logarithms. A polynomial is continuous everywhere on ℝ, a rational function is continuous wherever its denominator is non‑zero, and the trigonometric functions sin x and cos x are continuous on ℝ. The exponential function ex and the natural logarithm ln x are continuous on their respective domains (ℝ and (0, ∞)). These facts are frequently used without proof in JEE problems, but a student should be able to justify them by appealing to the limit laws and the standard limits discussed earlier.
3. Differentiability
3.1 Definition of the Derivative
The derivative of a function f at a point a, denoted f′(a) or df/dx|x=a, is defined as the limit limh→0 [f(a+h) − f(a)]/h, provided this limit exists. Geometrically, the derivative represents the slope of the tangent line to the curve y = f(x) at the point (a, f(a)). The existence of this limit implies that the function is locally well approximated by a linear function, and the error of this approximation tends to zero faster than h. The definition also makes clear that differentiability is a stronger condition than continuity, because the limit of the difference quotient requires the numerator f(a+h) − f(a) to tend to zero as h → 0, which is precisely the condition for continuity at a.
If the limit exists, we say f is differentiable at a. The derivative function f′(x) is then defined on the set of points where f is differentiable. In many JEE problems, the derivative is computed using first principles for simple functions such as x2, sin x, or ex to illustrate the process, after which standard differentiation rules are employed.
3.2 Differentiability Implies Continuity
If a function is differentiable at a point, it must be continuous there. The proof follows directly from the definition: f(a+h) − f(a) = [f(a+h) − f(a)]/h · h. Taking the limit as h → 0, the first factor tends to f′(a) (finite) and the second factor tends to 0, so the product tends to 0. Hence limh→0 f(a+h) = f(a), which is exactly continuity at a. The converse is false; a classic counterexample is f(x) = |x| at x = 0, which is continuous but not differentiable because the left‑hand and right‑hand difference quotients give −1 and +1 respectively. This distinction is frequently tested: students must be able to produce examples of continuous but non‑differentiable functions and explain why the derivative fails to exist.
3.3 Rules of Differentiation
Once the derivatives of elementary functions are known, the derivative of any combination can be obtained using a small set of rules. The sum rule states (f+g)′ = f′ + g′; the constant multiple rule gives (cf)′ = c f′; the product rule (Leibniz rule) is (fg)′ = f′g + fg′; the quotient rule is (f/g)′ = (f′g − fg′)/g2 for g ≠ 0; and the chain rule for composition reads (f∘g)′(x) = f′(g(x))·g′(x). These rules are derived from the limit definition and the algebra of limits, and they are indispensable for differentiating algebraic, trigonometric, exponential, and logarithmic functions efficiently. In the JEE context, the chain rule is especially important because many problems involve nested functions such as sin(ex2) or ln(tan x). Mastery of these rules, together with the standard derivatives d/dx (xn) = n xn−1, d/dx (sin x) = cos x, d/dx (cos x) = −sin x, d/dx (ex) = ex, d/dx (ln x) = 1/x, allows a student to differentiate virtually any function appearing in the syllabus.
3.4 Higher Order Derivatives
The second derivative f″(x) is the derivative of f′(x), and in general the nth derivative f(n)(x) is obtained by differentiating n times. Higher order derivatives provide information about the curvature and the rate of change of the slope. For instance, f″(x) > 0 indicates the graph is concave upward, while f″(x) < 0 indicates concave downward. In physics‑related problems, the second derivative of position with respect to time gives acceleration. The Leibniz rule for the nth derivative of a product, (fg)(n) = Σk=0n C(n,k) f(k) g(n−k), is a direct generalisation of the product rule and is occasionally required for series expansions or for solving differential equations. Students should be comfortable computing up to the third derivative of standard functions and recognising patterns such as the cyclic derivatives of sin x and cos x.
| Aspect | Rolle’s Theorem | Lagrange’s Mean Value Theorem |
|---|---|---|
| Hypotheses | f continuous on [a,b], differentiable on (a,b), f(a)=f(b) | f continuous on [a,b], differentiable on (a,b) |
| Conclusion | ∃ c∈(a,b) such that f′(c)=0 | ∃ c∈(a,b) such that f′(c) = (f(b)−f(a))/(b−a) |
| Geometric Meaning | Tangent parallel to x‑axis | Tangent parallel to chord joining (a,f(a)) and (b,f(b)) |
| Special Case | Lagrange with f(a)=f(b) | Generalisation of Rolle |
4. Mean Value Theorems
4.1 Rolle’s Theorem
Rolle’s theorem is a fundamental existence result that guarantees a stationary point for a differentiable function that takes equal values at the endpoints of a closed interval. Precisely, if f is continuous on [a, b], differentiable on (a, b), and f(a) = f(b), then there exists at least one c in (a, b) such that f′(c) = 0. The proof uses the Extreme Value Theorem to assert that f attains a maximum and a minimum on [a, b]; if both occur at the endpoints, the function is constant and the derivative is zero everywhere. Otherwise, an interior extremum exists, and Fermat’s theorem on stationary points gives f′(c) = 0. Rolle’s theorem is often the first step in proving the more general mean value theorems and is also used to establish the uniqueness of roots of equations: if a differentiable function has two distinct zeros, its derivative must vanish somewhere between them.
4.2 Lagrange’s Mean Value Theorem
Lagrange’s mean value theorem (often called simply the Mean Value Theorem) extends Rolle’s theorem by removing the requirement f(a) = f(b). It states that if f is continuous on [a, b] and differentiable on (a, b), then there exists a point c in (a, b) such that f′(c) = (f(b) − f(a))/(b − a). The right‑hand side is the slope of the secant line joining the endpoints of the graph. The theorem is proved by applying Rolle’s theorem to the auxiliary function g(x) = f(x) − [(f(b) − f(a))/(b − a)](x − a) − f(a), which satisfies g(a) = g(b) = 0. The conclusion f′(c) = (f(b) − f(a))/(b − a) has far‑reaching consequences: it provides a link between the average rate of change over an interval and the instantaneous rate of change at some interior point, and it underpins many inequalities and approximation results, including the proof of the Fundamental Theorem of Calculus.
4.3 Cauchy’s Mean Value Theorem
Cauchy’s mean value theorem generalises Lagrange’s theorem to two functions. If f and g are continuous on [a, b] and differentiable on (a, b), and g′(x) ≠ 0 for all x in (a, b), then there exists c in (a, b) such that (f(b) − f(a))/(g(b) − g(a)) = f′(c)/g′(c). This result is obtained by applying Rolle’s theorem to the function h(x) = f(x) − f(a) − [(f(b) − f(a))/(g(b) − g(a))](g(x) − g(a)). Cauchy’s theorem is the basis for L’Hôpital’s rule, because taking the limit as b → a yields the familiar form limx→a f(x)/g(x) = limx→a f′(x)/g′(x) under appropriate conditions. In JEE problems, Cauchy’s theorem sometimes appears in disguise when a ratio of differences is given and the student is asked to relate it to a ratio of derivatives.
4.4 Applications of the Mean Value Theorems
The mean value theorems are not merely theoretical; they provide powerful tools for estimating function values, proving inequalities, and establishing properties such as monotonicity. For example, if f′(x) > 0 on an interval, then by the Mean Value Theorem f is strictly increasing on that interval. Conversely, if f is increasing, its derivative cannot be negative anywhere it exists. The theorems also give error bounds for linear approximations: the difference between f(x) and its tangent line at a is exactly f′(c)(x−a) for some c between a and x. In competitive examinations, a typical question might ask to prove that sin x < x for x > 0 by applying the Mean Value Theorem to f(x) = sin x on [0, x] and using the fact that cos c < 1 for c in (0, x). Another common application is to show that a polynomial of degree n has at most n real roots by repeatedly applying Rolle’s theorem to its derivatives.
5. Problem Solving Strategies and Advanced Applications
5.1 Evaluating Indeterminate Forms Using L’Hôpital’s Rule
L’Hôpital’s rule is a direct corollary of Cauchy’s mean value theorem and provides a systematic method for resolving limits of the indeterminate forms 0/0 and ∞/∞. The rule states that if limx→a f(x) = limx→a g(x) = 0 or both limits are infinite, and the limit of the quotient of derivatives limx→a f′(x)/g′(x) exists (or is infinite), then limx→a f(x)/g(x) = limx→a f′(x)/g′(x). The rule can be applied repeatedly as long as the indeterminate form persists. However, it is essential to verify the hypotheses each time: the functions must be differentiable near a (except possibly at a), and g′(x) must not vanish in a punctured neighbourhood of a. In practice, many JEE limits that yield 0/0 can be solved more elegantly by algebraic manipulation and standard limits, but L’Hôpital’s rule remains a reliable fallback, especially when the functions involve exponentials, logarithms, or inverse trigonometric functions whose standard limits are less familiar.
5.2 Using Series Expansions for Limits
Taylor and Maclaurin series provide another powerful technique for evaluating limits, particularly when the indeterminate form involves higher‑order small quantities. The Maclaurin series of a function f about 0 is f(x) = f(0) + f′(0)x + f″(0)x2/2! + … . By substituting the series expansions of the numerator and denominator, the lowest‑order non‑zero terms often determine the limit. For example, limx→0 (ex − 1 − x)/x2 can be found by writing ex = 1 + x + x2/2 + x3/6 + …, which gives the numerator as x2/2 + higher order terms, so the limit is 1/2. This method is especially useful when repeated application of L’Hôpital’s rule becomes cumbersome. Students should memorise the Maclaurin series of ex, sin x, cos x, ln(1+x), and (1+x)α up to the x3 term, as these cover the vast majority of series‑based limit problems in the examination.
5.3 Continuity and Differentiability in Parametric and Implicit Functions
Many JEE problems present curves in parametric form x = φ(t), y = ψ(t) or implicitly as F(x, y) = 0. The derivative dy/dx for a parametric curve is given by (dy/dt)/(dx/dt) provided dx/dt ≠ 0. Continuity and differentiability of the parametric functions φ and ψ translate directly into continuity and differentiability of the curve, except at points where dx/dt = 0, which may correspond to vertical tangents or cusps. For an implicit function defined by F(x, y) = 0, the implicit differentiation formula dy/dx = −Fx/Fy holds wherever Fy ≠ 0, a consequence of the Implicit Function Theorem. Understanding these relationships allows a student to analyse the behaviour of curves that are not graphs of functions in the usual sense, such as the cycloid or the folium of Descartes, and to determine points of non‑differentiability like cusps or self‑intersections.
5.4 Common Pitfalls and How to Avoid Them
A frequent mistake is to assume that the existence of a limit guarantees continuity; the function value must also be checked. Another common error is applying L’Hôpital’s rule to forms that are not 0/0 or ∞/∞, such as 0·∞ or ∞−∞, without first converting them into a quotient. Students also sometimes forget that the chain rule requires the inner function to be differentiable at the relevant point and the outer function to be differentiable at the image of that point. In mean value theorem problems, it is crucial to verify the continuity and differentiability hypotheses on the exact interval given; overlooking a point where the derivative does not exist invalidates the conclusion.
Uniform Continuity and Its Distinction from Pointwise Continuity
Pointwise continuity at x₀ means: for every ε > 0 there exists δ > 0 (depending on ε and x₀) such that |f(x) − f(x₀)| < ε whenever |x − x₀| < δ. Uniform continuity strengthens this by requiring a single δ for all points in the domain: for every ε > 0 there exists δ > 0 such that for all x, y in the domain, |x − y| < δ implies |f(x) − f(y)| < ε. The distinction is crucial on unbounded intervals or intervals that are not closed. For example, f(x) = 1/x is continuous on (0,1] but not uniformly continuous because as x → 0⁺ the required δ shrinks to zero. In contrast, the Heine–Cantor theorem guarantees that every continuous function on a closed, bounded interval [a,b] is uniformly continuous. This theorem is frequently used to justify term‑by‑term integration of series and to prove the existence of the Riemann integral for continuous functions.
A practical criterion for uniform continuity on ℝ is the Lipschitz condition: if |f(x) − f(y)| ≤ L|x − y| for all x, y, then f is uniformly continuous. Differentiable functions with bounded derivative satisfy this by the Mean Value Theorem. However, uniform continuity does not imply differentiability; |x| is uniformly continuous on ℝ but not differentiable at 0. Conversely, a function can be differentiable everywhere with unbounded derivative (e.g., f(x) = x² on ℝ) and fail to be uniformly continuous. Understanding these nuances helps decide whether a limit of a sequence of functions can be interchanged with integration or differentiation.
Sequential Characterisation
A function f is uniformly continuous on a set D iff for any two sequences (xₙ) and (yₙ) in D with |xₙ − yₙ| → 0, we have |f(xₙ) − f(yₙ)| → 0. This sequential test is often easier to apply in proofs because it avoids explicit ε‑δ manipulation. For instance, to show that f(x) = sin(x²) is not uniformly continuous on ℝ, choose xₙ = √(2πn) and yₙ = √(2πn + π/2). Then |xₙ − yₙ| → 0 but |sin(xₙ²) − sin(yₙ²)| = 1 does not tend to 0.
Darboux's Theorem and the Intermediate Value Property of Derivatives
While the Intermediate Value Theorem applies to continuous functions, Darboux's theorem states that derivatives also possess the intermediate value property, even if they are not continuous. Formally, if f is differentiable on an interval I and a, b ∈ I with a < b, then for any L between f′(a) and f′(b) there exists c ∈ (a, b) such that f′(c) = L. The proof applies the Mean Value Theorem to g(x) = f(x) − Lx. A striking consequence: a derivative cannot have a simple jump discontinuity. If f′ exists everywhere on an interval, its discontinuities must be oscillatory (e.g., f(x) = x² sin(1/x) for x ≠ 0, f(0) = 0, whose derivative at 0 is 0 but oscillates near 0). Darboux's theorem also implies that if a derivative takes both positive and negative values on an interval, it must vanish somewhere — a fact used to prove the existence of critical points without assuming continuity of the derivative.
An important corollary: if f′ exists on [a,b] and f′(a) < 0 < f′(b), then there is at least one c ∈ (a, b) with f′(c) = 0. This follows directly from the intermediate value property for derivatives; it does not require f′ to be continuous, only that it exists everywhere. The theorem also underpins the proof that a function with derivative identically zero on an interval is constant. In competitive problems, Darboux's theorem is often invoked to rule out certain discontinuities in a derivative or to guarantee a point where the derivative equals a prescribed value.
Taylor's Theorem with Remainder Forms and Error Estimation
Taylor's theorem generalises the Mean Value Theorem by approximating a function with a polynomial plus a remainder. If f has derivatives up to order n+1 on an interval containing a and x, then for each x there exists ξ between a and x such that
f(x) = f(a) + f′(a)(x−a) + f″(a)(x−a)²/2! + … + f⁽ⁿ⁾(a)(x−a)ⁿ/n! + Rₙ(x),
where the remainder Rₙ(x) has several forms. The Lagrange form is Rₙ(x) = f⁽ⁿ⁺¹⁾(ξ)(x−a)ⁿ⁺¹/(n+1)!. The Cauchy form is Rₙ(x) = f⁽ⁿ⁺¹⁾(ξ)(x−ξ)ⁿ(x−a)/n!. The integral form, valid when f⁽ⁿ⁺¹⁾ is integrable, is Rₙ(x) = ∫ax f⁽ⁿ⁺¹⁾(t)(x−t)ⁿ/n! dt. Each form suits different estimation tasks. The Lagrange form gives a simple bound if a uniform bound M for |f⁽ⁿ⁺¹⁾| is known: |Rₙ(x)| ≤ M|x−a|ⁿ⁺¹/(n+1)!. This underpins error estimates in numerical methods and convergence proofs for Maclaurin series of ex, sin x, cos x.
In examinations, a typical question asks to approximate √1.01 using a second‑order Taylor polynomial of √x about a = 1 and to bound the error. The Lagrange remainder with n = 2 involves the third derivative of √x, which is (3/8)x⁻⁵ᐟ². On [1, 1.01] the maximum occurs at x = 1, giving M = 3/8. The error bound is then (3/8)(0.01)³/6 ≈ 6.25×10⁻⁸. Another common task is to prove inequalities like sin x < x − x³/6 + x⁵/120 for sufficiently small x > 0 by examining the sign of the remainder. Mastery of the three remainder forms allows a student to choose the one yielding the tightest bound.
Maclaurin Series and Radius of Convergence
The Maclaurin series is the Taylor series about a = 0. For ex, sin x, cos x, the series converge for all real x because the remainder tends to zero as n → ∞. For ln(1+x) and (1+x)α, the series converge only for |x| < 1 (and at x = 1 under certain conditions on α). The radius of convergence can be found via the ratio test on the coefficients. In competitive problems, recognising a function's series up to x³ or x⁴ often suffices to evaluate limits or approximate integrals. For example, ∫00.1 sin(t²) dt can be approximated by integrating the series t² − t⁶/6 + … term by term, with the error controlled by the first omitted term.
Higher‑Order Derivatives: Leibniz Rule and Faà di Bruno's Formula
The nth derivative of a product u and v is given by the Leibniz rule:
(uv)⁽ⁿ⁾ = ∑k=0n C(n, k) u⁽ᵏ⁾ v⁽ⁿ⁻ᵏ⁾,
where C(n, k) = n!/(k!(n−k!)). This is proved by induction using the product rule and Pascal's identity. It is extremely useful when one factor is a polynomial (so its higher derivatives vanish) or when both factors have simple derivative patterns (e.g., ex sin x). For instance, the 5th derivative of x² ex requires only terms with k = 0,1,2 because u = x² has zero derivatives for k ≥ 3, reducing the computation to a short sum.
For the nth derivative of a composite function f(g(x)), Faà di Bruno's formula generalises the chain rule. It expresses the derivative as a sum over partitions of n:
dⁿ/dxⁿ f(g(x)) = ∑ n!/(m₁! m₂! … mₙ!) f⁽ᵐ⁾(g(x)) ∏j=1n (g⁽ʲ⁾(x)/j!)^mⱼ,
where the sum runs over all n-tuples of non‑negative integers (m₁,…,\mₙ) satisfying 1·m₁ + 2·m₂ + … + n·mₙ = n, and m = m₁ + … + mₙ. The full formula is rarely needed; the first few cases are standard: the second derivative is f″(g)(g′)² + f′(g)g″, the third adds terms involving g‴ and products of g′ and g″. In examinations, problems often ask for the nth derivative of a specific composite function like sin(x²) or ln(1+x²) at x = 0, which can be tackled by combining series expansions with the Leibniz rule rather than by direct application of Faà di Bruno. Nevertheless, knowing the formula's structure helps recognise patterns and prove identities involving higher derivatives.
Applications to Differential Equations
Higher‑order derivatives appear naturally in linear differential equations with constant coefficients. The characteristic polynomial method relies on the fact that the nth derivative of eλx is λⁿ eλx. For equations with variable coefficients, the Leibniz rule is used to compute derivatives of products that arise in reduction of order or in finding integrating factors. For example, to solve y″ − 2xy′ + 2y = 0, one may look for a solution of the form y = u(x)v(x) and use the Leibniz rule to express the derivatives of the product. Mastery of these rules is both a computational skill and a gateway to broader problem-solving.
- Assuming uniform continuity on an unbounded interval just because the function is continuous.
- Applying the Intermediate Value Theorem to a derivative without verifying that the derivative exists at every point of the interval.
- Using the Lagrange remainder without checking that the (n+1)th derivative is bounded on the interval between a and x.
- Forgetting that the Leibniz rule sum stops when one factor's derivatives become zero, leading to unnecessary terms.
- Misapplying Faà di Bruno's formula by omitting the multinomial coefficient or using incorrect partition conditions.
- Confusing the Cauchy form of the remainder with the Lagrange form when estimating errors.
- Neglecting to verify the hypotheses of Darboux's theorem (differentiability on the whole interval) before concluding the existence of a zero of the derivative.