When using a Taylor polynomial of degree n centered at c to approximate the value of a function f at x, there is an error because the polynomial does not exactly mimic the function (unless, of course, f is a polynomial of degree less than or equal to n). We can bound this error using the Lagrange remainder (or Lagrange error bound). The remainder is:
where M is the maximum of the absolute value of the (n + 1)th derivative of f on the interval from x to c. The error is bounded by this remainder (i.e., the absolute value of the error is less than or equal to R). Note that R depends on how far x is away from c, how big n is, and on the characteristics of f.
|See About the calculus applets for operating instructions.
The applet shows the Taylor polynomial with n = 3, c = 0 and x = 1 for f (x) = ex. To compute the Lagrange remainder we need to know the maximum of the absolute value of the 4th derivative of f on the interval from 0 to 1. Since the 4th derivative of ex is just ex, and this is a monotonically increasing function, the maximum value occurs at x = 1 and is just e. So:
Note in the applet that the actual error is about 0.052. The Lagrange remainder is a bound on the error, not the actual error itself. It just says that the error, whatever it is, will be less than the Lagrange remainder.
While you are here..
... I have a small favor to ask. Over the years we have used advertising to support the site so it can remain free for everyone.
However, advertising revenue is falling and I have always hated the ads. So, would you go to Patreon and become a patron of the site?
When we reach the goal I will remove all advertising from the site.
It only takes a minute and any amount would be greatly appreciated.
Thank you for considering it! – John Page
Become a patron of the site at patreon.com/mathopenref
Other 'Sequences and Series' topics
Derived from the work of Thomas S. Downey under a Creative Commons Attribution 3.0 License.