∿ExploreSims
ExploreSims / Mathematics / Calculus / Newton's Method & Root Finding
Premium experiment • numerical root finding

MATHEMATICS • CALCULUS

Newton's Method & Root Finding

Follow tangent-line updates from an initial guess and see how quickly different nonlinear equations converge toward a root.

● ITERATIVE ROOT FINDINGTangent-line updates
Root estimate
Residual
Iterations
Status

Reasoning inspector

live model state
Newton estimate—
Residual |f(x)|—
Iterations used—
Last derivative—

Local slope becomes a root estimate

Newton's method replaces the function locally with its tangent line, then uses that line's x-intercept as the next approximation.

LEARN WITH THE SIMULATION

Understand Newton's Method & Root Finding by experimenting.

START HERE

Newton's method is an iterative root-finding algorithm that turns local slope information into a sequence of guesses. At each step it draws the tangent line to the current point, follows that tangent to the x-axis, and uses the intercept as the next estimate. The method is powerful because convergence can be extremely fast near a well-behaved root, but its behavior also exposes an important numerical lesson: the derivative, the starting point and the shape of the function all matter.

BUILD THE IDEA

What you should understand first

  • Apply x(n+1) = x(n) - f(x(n))/f'(x(n)) and connect every algebraic update to the tangent shown on the graph.
  • Judge convergence using both the changing root estimate and the residual |f(x)| rather than trusting a visually plausible x-value.
  • Compare starting points to see rapid convergence, large jumps, slow progress and failure near nearly horizontal tangents.
  • Explain why Newton's method is locally powerful but not globally guaranteed to find a root.

TRY IT YOURSELF

How much can the starting guess change the entire Newton iteration?

  1. Choose the cubic function, set a starting value close to a visible root and run enough iterations to reach a very small residual.
  2. Move the starting value farther away while keeping the same function and iteration budget. Watch the tangent intercepts, not just the final number.
  3. Try a starting value where the graph is comparatively flat. Compare step size, residual and whether the sequence moves toward or away from a root.
  4. Repeat with another nonlinear function and decide whether the same starting-point intuition still works.
Read the visual:

Near a simple root with a useful derivative, Newton steps usually shrink rapidly and the residual can fall dramatically. Farther away, the tangent line can point to a distant part of the graph; near a small derivative, dividing by f'(x) can produce a very large step. The method responds to local geometry, not to the nearest root in any global sense.

WHY IT WORKS 01

What Newton's method is really doing

The update is not a mysterious formula: it comes from replacing the nonlinear function near x_n with its tangent line. The tangent has slope f'(x_n), so its x-intercept is x_n - f(x_n)/f'(x_n). Newton's method repeatedly rebuilds that local linear approximation. This is why the derivative controls both direction and step size, and why the method becomes especially accurate when the current guess is already near a simple root.

WHY IT WORKS 02

Why the starting point matters

Newton's method has no global map of the function. It only sees the value and derivative at the current guess. Two nearby starting points can therefore enter different basins of attraction and converge to different roots. A starting point that lands near a flat tangent can jump far away, while one already inside a favorable basin may converge in only a few iterations.

WHY IT WORKS 03

When Newton's method fails

Failure can occur when f'(x) is zero or very small, when iterations cycle between values, when the sequence diverges, or when the function is not well behaved in the region being explored. In practical numerical software, Newton's method is often combined with safeguards such as bracketing, damping or line searches. This simulation intentionally exposes the unsafeguarded method so those failure modes are visible rather than hidden.

DO NOT MIX THESE UP

Common misconception

Newton's method is not a calculator that is guaranteed to return the nearest root, and a small change in the initial guess does not always produce a small change in the iteration path. A small residual is stronger evidence of a root than a stable-looking x-value. A zero or nearly zero derivative is not merely a display problem: it changes the mathematics of the update and can make the next step undefined or extremely large.

PRACTICE CHALLENGE

Can you predict a bad Newton step before pressing the next iteration?

  • Choose a function and place the initial guess where the tangent looks nearly horizontal.
  • Before advancing, predict whether the next x-value will move a small distance or make a large jump.
  • Use the tangent geometry and the size of f'(x) to justify your prediction, then run the step and compare.
Reveal the reasoning

A nearly horizontal tangent means |f'(x)| is small. Because Newton divides f(x) by f'(x), the correction term can become very large unless f(x) is also very small. Geometrically, a shallow tangent reaches the x-axis far from the current point, so a large jump is exactly what the formula predicts.

MODEL BOUNDARIES

What this simulation assumes

Every simulation simplifies reality or a mathematical system. These are the assumptions and limits to keep in mind while interpreting the results.

Model assumptions

  • Newton updates use exact analytic derivatives for three predefined functions.
  • The displayed iteration starts from a user-selected scalar initial guess and uses ordinary double-precision arithmetic.
  • Convergence is classified when the final absolute residual is below 1e-8.

Model limitations

  • The lab does not implement bracketing, line search, derivative-free fallbacks or complex roots.
  • Some initial guesses can diverge, cycle or encounter a nearly zero derivative; those behaviors are properties of Newton's method rather than numerical guarantees.

CONNECT IT

Where this matters

  • Solving nonlinear engineering and physics equations where closed-form roots are unavailable
  • Implicit equations inside numerical models, calibration routines and scientific computing
  • Optimization algorithms, where Newton-style curvature updates generalize the same local-approximation idea
  • Numerical analysis for studying convergence, stability and sensitivity to initial conditions

WORDS THAT MATTER

Key vocabulary

Root
A value of x for which f(x) = 0.
Newton iteration
The update x(n+1) = x(n) - f(x(n))/f'(x(n)), obtained from the current tangent-line intercept.
Residual
The magnitude |f(x)| at the current estimate; a direct measure of how closely the equation is being satisfied.
Derivative
The local slope f'(x), which determines the tangent direction and scales the Newton correction.
Basin of attraction
A set of starting values whose Newton iterations converge to the same root.