Jump to content

Banach fixed-point theorem

From Wikiversity
Subject classification: this is a mathematics resource.
Educational level: this is a tertiary (university) resource.

Banach fixed-point theorem. Let be a non-empty complete metric space i.e. the space with a metric where every Cauchy sequence i.e. such that its elements are all arbitrarily closely bunched in the sense of the metric if their indexes are sufficiently high is convergent with a contraction mapping i.e if there exists such that

i.e. that the metric between the images of any of two elements is contracted or scaled by the factor less than 1 with respect to the metric between elements themself for all . Then there is only one unique fixed point fixed-point for in (i.e. or the element such that it image is the same that the element). Furthermore, can be found by the infinite nesting of T or the infinite superposition applied on any element of the space has the fixed point limit: start with an arbitrary element and define a sequence by for Then .

Example of numerical applications - calculating high accuracy π

[edit | edit source]

Banach theorem allows for example fast and accurate calculation of the π number using the trigonometric functions which numerically are the power Taylor series.

Because and the π is the fixed point of for example the function

i.e.

and also the function is around π the contraction mapping from the obvious reasons because its derivative at π vanishes at the bending point therefore π can be obtained from the infinite superposition for example for the argument value 3:

Already the triple superposition of this function at gives π with accuracy to 33 digits:

.