Applications of Differentiation — Topic 06

Newton's Method

The most powerful and widely used root-finding algorithm in mathematics. Starting from an initial guess, Newton's Method repeatedly replaces that guess with the x-intercept of the tangent line at the current point, and under the right conditions, the number of correct decimal places roughly doubles with every single iteration.

1Core Formula
8Worked Examples
5Failure Modes
10Quiz Questions
Share this page
On This Page
  1. The Root-Finding Problem
  2. Deriving the Iteration Formula
  3. Geometric Interpretation, Step by Step
  4. Convergence and Quadratic Speed
  5. The Algorithm in Practice
  6. Worked Examples
  7. When Newton's Method Fails
  8. Common Mistakes
  9. 10-Question Quiz

§ 01The Root-Finding Problem

A root (or zero) of a function f is a value x = r such that f(r) = 0. Finding roots is one of the oldest and most important problems in all of mathematics, equations of the form f(x) = 0 arise constantly in physics, engineering, economics, and computer science.

For simple functions, roots can be found exactly using algebra. The quadratic formula gives the roots of any quadratic polynomial. But for equations like x⁵ − 3x + 1 = 0, or cos x = x, or ex = 3x², there is no general algebraic formula. We need a different approach, a numerical method that produces increasingly accurate approximations to the root.

The oldest numerical approach is the bisection method: if f(a) and f(b) have opposite signs, the Intermediate Value Theorem guarantees a root in (a, b), so you repeatedly halve the interval to zero in on it. Bisection is reliable but slow: it adds roughly one correct binary digit per iteration, which means about 3.3 iterations per additional decimal place.

Newton's Method (also called the Newton–Raphson method) is dramatically faster under the right conditions. It uses not just the value of f but also the derivative f′ (the tangent line) to make a much more intelligent estimate of where the root lies. When it works, it is roughly quadratically convergent: the number of correct decimal places doubles with each iteration. That means two correct digits become four, then eight, then sixteen. In practice, six or seven iterations typically gives you fifteen-digit machine precision.

Historical Note Isaac Newton described the core idea in his 1669 work De analysi, though his formulation was different from the modern one. Joseph Raphson gave the cleaner iterative version in 1690. The method was further refined by Thomas Simpson in 1740 into the form we use today. It is one of the earliest algorithms to be analysed in terms of speed of convergence, and remains the standard root-finding method in scientific computing.

§ 02Deriving the Iteration Formula

The derivation is based on a single geometric insight: replace the curve with its tangent line, find where that line crosses the x-axis, and use that crossing as the next approximation.

Suppose we have a current approximation xn to the root. The tangent line to y = f(x) at the point (xn, f(xn)) has the equation:

Tangent line at xn
y − f(xn) = f′(xn)(x − xn)

This is just the point–slope form of a line through (xn, f(xn)) with slope f′(xn).

To find where this tangent line crosses the x-axis, set y = 0 and solve for x:

Solving for the x-intercept of the tangent
0 − f(xn) = f′(xn)(x − xn) x − xn = −f(xn) / f′(xn) x = xn − f(xn) / f′(xn)

This x-intercept is our improved approximation xn+1.

This gives the Newton–Raphson iteration formula:

Newton's Method — The Iteration Formula
xn+1 = xn − f(xn) / f′(xn)
Starting from an initial guess x0, each new approximation is obtained
by subtracting the ratio of the function value to the derivative at the current point.
Repeat until |xn+1 − xn| or |f(xn+1)| is smaller than your desired tolerance.

The formula requires that f′(xn) ≠ 0 at every step. If the derivative is zero at some iterate, the tangent line is horizontal and has no x-intercept, the method breaks down at that point. This is one of the key failure modes we examine in § 07.

What the Formula Means Intuitively The correction term −f(xn)/f′(xn) is a Newton step. It asks: "If f were perfectly linear with slope f′(xn), how far would I need to move from xn to reach f = 0?" The answer is −f(xn)/f′(xn). For a truly linear function this gives the exact root in one step. For nonlinear functions it gives only an approximation — but a much better one than xn, because the tangent line is locally an excellent approximation to the curve.

§ 03Geometric Interpretation — Step by Step

Every iteration of Newton's Method has an identical geometric description: draw the tangent line at the current point, find where it hits the x-axis, and move there. The diagram below shows three successive iterations converging rapidly to the root.

Newton's Method — Three Iterations Converging to Root r
x y r x0 x1 x2 f(x₀) tangent tangent root r x₀ → x₁
Gold: starting point x₀ and its tangent line. Red: x₁ and its tangent. Grey: x₂, very close to the root r (teal dot). Each tangent line intersects the x-axis at the next iterate — the steps get dramatically smaller as the root is approached.

Notice how the distances between successive iterates shrink: x₀ to x₁ is a large step; x₁ to x₂ is a smaller step; x₂ to x₃ would be tiny. This accelerating convergence is the signature of Newton's Method and is what distinguishes it from slower algorithms like bisection.

The geometric picture also reveals when the method might have trouble: if the tangent line at xₙ is nearly horizontal (f′ ≈ 0), its x-intercept will be far away from xₙ, potentially overshooting the root entirely. If the curve has a lot of curvature near the starting guess, the tangent line may not be a good approximation to the curve over a large step. These geometric observations motivate the failure analysis in § 07.

§ 04Convergence and Quadratic Speed

The most remarkable feature of Newton's Method is how fast it converges when it works. This speed is described precisely by the concept of order of convergence.

Linear (First-Order) Convergence
Bisection Method
|eₙ₊₁| ≈ C · |eₙ|
The error is multiplied by a constant C < 1 each step. If C = 0.5 (bisection), each step halves the error — about 3.3 steps per new decimal digit. To gain 15 decimal digits from 1 requires roughly 50 steps.
Quadratic (Second-Order) Convergence
Newton's Method
|eₙ₊₁| ≈ C · |eₙ|²
The error is roughly squared each step (multiplied by a constant). If |e₀| = 0.1, then |e₁| ≈ 0.01, |e₂| ≈ 0.0001, |e₃| ≈ 10⁻⁸, |e₄| ≈ 10⁻¹⁶. Four steps from 1 correct digit to 16. This is the "doubling of digits" behaviour.

Where Does Quadratic Convergence Come From?

Let r be the exact root (f(r) = 0) and let eₙ = xₙ − r be the error at step n. Expand f(xₙ) and f′(xₙ) in Taylor series around r:

Taylor expansion of the error
f(xₙ) = f(r) + f′(r)eₙ + ½f″(r)eₙ² + ···         = f′(r)eₙ + ½f″(r)eₙ² + ···    [since f(r)=0] f′(xₙ) = f′(r) + f″(r)eₙ + ···
Newton step error analysis
xₙ₊₁ = xₙ − f(xₙ)/f′(xₙ) eₙ₊₁ = xₙ₊₁ − r = eₙ − f(xₙ)/f′(xₙ) eₙ₊₁ ≈ eₙ − [f′(r)eₙ + ½f″(r)eₙ²] / f′(r) eₙ₊₁ ≈ eₙ − eₙ − [f″(r)/(2f′(r))]eₙ² eₙ₊₁ ≈ − [f″(r)/(2f′(r))] · eₙ²

The eₙ terms cancel exactly! What remains is proportional to eₙ². This is quadratic convergence.

The convergence constant is C = |f″(r)|/(2|f′(r)|). For convergence, we need f′(r) ≠ 0 (the root must be simple, not a repeated root). If the root is a double root (f(r) = f′(r) = 0), Newton's Method degrades to linear convergence with rate C = 1/2.

Simple vs Repeated Roots At a simple root (f(r) = 0 but f′(r) ≠ 0), Newton's Method converges quadratically. At a double root (f(r) = f′(r) = 0 but f″(r) ≠ 0), convergence is only linear with rate ½ — Newton's Method is effectively no better than bisection. For a root of multiplicity m, the convergence rate is (m−1)/m, which approaches 1 (very slow) for high-multiplicity roots. The fix is a modified iteration: xₙ₊₁ = xₙ − m·f(xₙ)/f′(xₙ), which restores quadratic convergence.

§ 05The Algorithm in Practice

Here is Newton's Method laid out as a step-by-step algorithm you can follow by hand or implement in code.

Newton's Method Algorithm

Input: function f, derivative f′, initial guess x₀, tolerance ε (e.g. ε = 10⁻⁶), maximum iterations N.

1. Set n = 0.

2. Compute xn+1 = xn − f(xn)/f′(xn). Check that f′(xn) ≠ 0 first.

3. If |xn+1 − xn| < ε or |f(xn+1)| < ε, stop, convergence achieved.

4. Set n = n + 1. If n < N, go to step 2. Otherwise stop: did not converge.

Output: xn+1 as the approximate root.

Choosing a Good Starting Point

The starting guess x₀ is the single most important decision. Newton's Method is guaranteed to converge only when x₀ is "sufficiently close" to the root, a condition that is hard to quantify precisely in advance but can be guided by the following principles:

1. Plot f first. Even a rough sketch identifies approximately where f crosses the x-axis. Use this region to anchor x₀. The Intermediate Value Theorem confirms a root in (a, b) whenever f(a) and f(b) have opposite signs.

2. Use the sign of f. Bracket the root by finding a and b with f(a) < 0 and f(b) > 0. Start x₀ anywhere in (a, b), or use the midpoint as a safe starting value.

3. Avoid inflection points and local extrema. If x₀ happens to be near a local maximum or minimum of f, the derivative f′(x₀) will be small, making the correction term huge. Start away from stationary points of f.

Stopping Criterion: Two Common Choices You can stop Newton's Method when either (a) |xₙ₊₁ − xₙ| < ε — the successive approximations are close together — or (b) |f(xₙ₊₁)| < ε — the function value is nearly zero. Condition (a) measures how much x is changing; condition (b) measures how close to zero f is. In general, use both: stop when either is satisfied. Near a simple root with f′ not too small, the two criteria give similar results. Near a double root or near a horizontal asymptote of f, condition (a) may be met while (b) is not.
♦   ♦   ♦

§ 06Worked Examples

Example 01 Find a root of f(x) = x³ − 2x − 5  starting at x₀ = 2

This is Newton's own original example from 1669. The root is r ≈ 2.09455148…

Setup
f(x) = x³ − 2x − 5,    f′(x) = 3x² − 2

Formula: xn+1 = xn − (xn³ − 2xn − 5)/(3xn² − 2)

nxₙf(xₙ)f′(xₙ)xₙ₊₁
02.000000000−1.00000010.0000002.100000000
12.1000000000.06100011.2300002.094568121
22.0945681210.00013011.1539372.094551482
32.0945514820.00000011.1535722.094551482
Result
Root ≈ 2.094551482. Correct to 9 decimal places in just 3 iterations from x₀ = 2.
Example 02 Compute √2 using Newton's Method
Setup

√2 is the positive root of f(x) = x² − 2. So f′(x) = 2x.

xn+1 = xn − (xn² − 2)/(2xn) = (xn + 2/xn) / 2

This is the Babylonian method, one of the oldest known algorithms, now seen as a special case of Newton's Method. Start at x₀ = 1.

nxₙError |xₙ − √2|
01.0000000000000.414214
11.5000000000000.085786
21.4166666666670.002453
31.4142156862750.0000021
41.4142135623751.6 × 10⁻¹²
Observation
Digits double with each step: 1 → 2 → 4 → 8 → 12. This is quadratic convergence in action. √2 = 1.41421356237…
Example 03 Solve cos x = x (fixed-point equation), x₀ = 1
Rearrange

Write f(x) = cos x − x = 0. Then f′(x) = −sin x − 1.

xn+1 = xn − (cos xn − xn) / (−sin xn − 1)
nxₙf(xₙ) = cos xₙ − xₙ
01.000000−0.459698
10.7503640.018368
20.7391130.000033
30.739085< 10⁻¹⁰
The Dottie Number
x ≈ 0.739085133 (the Dottie number — the unique fixed point of cos). 3 iterations from x₀ = 1 gives 10-digit accuracy.
Example 04 Compute ∛7 using Newton's Method starting at x₀ = 2
Setup

∛7 is the positive root of f(x) = x³ − 7. So f′(x) = 3x².

xn+1 = xn − (xn³ − 7)/(3xn²) = (2xn/3) + 7/(3xn²)
nxₙxₙ³
02.0000000008.000
11.9166666677.036
21.9129391667.000064
31.912931183≈ 7.000000000
Result
∛7 ≈ 1.91293118. Three iterations give 9 correct decimal places from x₀ = 2.
Example 05 Find a positive root of f(x) = eˣ − 3x = 0, starting at x₀ = 1.5
Setup

f′(x) = eˣ − 3. Note: f has two roots (one near x ≈ 0.62, one near x ≈ 1.51). Starting at x₀ = 1.5 finds the larger root.

xn+1 = xn − (e^{x_n} − 3xn)/(e^{x_n} − 3)
nxₙf(xₙ)
01.500000−0.018
11.512134−0.000005
21.512138≈ 0
Result
Root ≈ 1.51213779. Two iterations suffice because x₀ was already very close.
Example 06 Why Newton's Method finds roots of linear functions in exactly one step
Claim

For f(x) = ax + b (a linear function), one Newton step from any x₀ gives the exact root in one iteration.

Proof
x1 = x0 − f(x0)/f′(x0) = x0 − (ax0+b)/a = x0 − x0 − b/a = −b/a

The exact root of ax + b = 0 is x = −b/a. Newton reaches it immediately because a tangent to a line is the line itself.

Insight
Newton's Method is exact for linear functions and very fast for functions that are "nearly linear" near their roots. The more curved the function near the root, the more iterations are needed.
Example 07 Find the root of f(x) = x¹/³ — demonstrating divergence from an infinite derivative at the root
Analysis

f(x) = x1/3 has a root at x = 0. But f′(x) = (1/3)x−2/3, which is undefined at x = 0. The root at x = 0 is not differentiable: Newton's Method cannot be applied directly at the root.

What happens
xn+1 = xn − xn1/3 / [(1/3)xn−2/3] = xn − 3xn = −2xn

The iterates oscillate: x₀ = 1 → x₁ = −2 → x₂ = 4 → x₃ = −8 → · · · Divergence!

Lesson
Newton's Method requires f to be differentiable at the root with f′(r) ≠ 0. When the root is not differentiable or f′(r) = 0, the method breaks down or converges slowly.
Example 08 f(x) = x² − 5: find both roots using different starting points
Setup

f′(x) = 2x. The roots are ±√5 ≈ ±2.2360679…

xn+1 = xn − (xn² − 5)/(2xn) = (xn + 5/xn)/2
x₀ = 2 → +√5

x₁ = (2 + 5/2)/2 = 2.25    x₂ = (2.25 + 5/2.25)/2 ≈ 2.23611    x₃ ≈ 2.23607 ✓

x₀ = −2 → −√5

x₁ = (−2 + 5/(−2))/2 = −2.25    x₂ ≈ −2.23611    x₃ ≈ −2.23607 ✓

Key Point
The starting point determines which root Newton's Method converges to. A positive x₀ finds +√5; a negative x₀ finds −√5. Starting at x₀ = 0 causes division by zero (f′(0) = 0) — an immediate failure.

§ 07When Newton's Method Fails

Newton's Method is not guaranteed to converge for every function and starting point. Understanding the failure modes is essential for using the method safely.

Failure 1
f′(xₙ) = 0 at some iterate
If the derivative is zero at any point in the iteration, the formula xn+1 = xn − f(xn)/f′(xn) involves division by zero and is undefined. This happens if the iteration lands exactly on a critical point of f.

Fix: Restart with a different x₀, or apply a small perturbation.
Failure 2
Overshooting — Cycling Between Two Points
If the tangent at xₙ sends the iteration to a region where the tangent sends it back, the method can cycle indefinitely between two values without converging. This is common for functions with significant curvature near the root.

Classic example: f(x) = x³ − 2x with x₀ = 1 cycles between 1 and −1.
Failure 3
Divergence — Iterates Escape to Infinity
If f(xn)/f′(xn) is consistently larger than the distance to the root, the iterates can grow without bound. This occurs when x₀ is far from the root in a region where f is very flat (|f′| ≈ 0).

Example: f(x) = arctan x with x₀ large: the tangent is nearly horizontal, the correction is huge.
Failure 4
Convergence to the Wrong Root
When f has multiple roots, Newton's Method converges to the root nearest the starting point — which may not be the root you want. The "basin of attraction" of each root is the set of starting points that converge to it, and for complicated functions this boundary can be fractal.

Fix: Bracket the desired root before starting, or test multiple starting points.
Failure 5
Slow Convergence at Repeated Roots
At a root of multiplicity m > 1, Newton's Method converges only linearly, not quadratically. The rate is (m−1)/m per step — for a double root this is ½, barely better than bisection.

Fix: Use the modified formula xₙ₊₁ = xₙ − m·f(xₙ)/f′(xₙ), which restores quadratic convergence when m is known.
Failure Mode 2 — Cycling Between Two Points
x y root r=0 x0 x0 x1 cycles forever — never reaches root
Tangent at x₀ (gold) hits the axis at x₁ on the opposite side. Tangent at x₁ (red) sends the iteration straight back to x₀. The method oscillates indefinitely and never converges to the root at the origin.

§ 08Common Mistakes

Mistake 1
Using the wrong derivative
Newton's formula requires f′(xₙ) — the derivative of f, evaluated at xₙ. A common error is differentiating f incorrectly (e.g. missing the chain rule) or evaluating f′ at the wrong point. Always write f′(x) explicitly before substituting, and double-check by differentiating again if unsure.
Mistake 2
Stopping too early
Declaring convergence after one or two iterations without checking the stopping criterion. Always compute |xₙ₊₁ − xₙ| or |f(xₙ₊₁)| and compare to the required tolerance ε. In exam questions, carry enough decimal places throughout — premature rounding corrupts later iterations.
Mistake 3
Not checking that x₀ gives a valid step
Before iterating, check f′(x₀) ≠ 0. If f′(x₀) = 0, the very first step is undefined. Similarly, if x₀ happens to be a critical point of f (a local extremum), the tangent is horizontal and the iteration fails immediately.
Mistake 4
Confusing f(x) = 0 with x = g(x)
Newton's Method solves f(x) = 0. If your equation is written as h(x) = k(x), first rearrange to f(x) = h(x) − k(x) = 0, then apply Newton's iteration to this f. Do not try to apply Newton's Method directly to the original two-sided form.

§ 0910-Question Quiz

Questions cover the formula, carrying out iterations, convergence properties, and identifying failure modes.

§ 10Continue the Applications Series

Cookie Settings