Newton's Method

A brief description

The applet below illustrates geometrically the approximation method known as Newton's method. Given a function f, Newton's method attempts to locate an input x such that f(x) = 0. Newton's method is an iterative approximation scheme: given an estimate, it seeks to find another (hopefully better) estimate. The method is then iterated, providing a sequence of estimates, each one generated by the same method from the previous estimate.

The basic idea is that the function might be zero close to the point where a tangent line to the function intersects the horizontal axis. So we use the intersection of the horizontal axis with the tangent line (at the current estimate) as our next estimate. As the examples provided here show, Newton's method sometimes works very efficiently, but other times it fails to work at all. More sophisticated methods generally combine Newton's method with other methods (like the biject method) to compensate for its deficiencies while taking advantage of its strengths.

Newton's iteration formula (the algebraic equation that shows how to get a new estimate from a current estimate) is easily derived from simple facts about slopes of lines and the geometric meaning of the derivative. It can also be found in most calculus text books.

Usage Notes

You can change the viewing rectangle in several ways:


This applet was written by Randall Pruim and makes use of the Java Components for Mathematics, written by David Eck. Last Modified: Friday, 16-Nov-2001 08:47:45 EST