View source for Simple continued fraction

{{Short description|Number represented as 1+a1/(1+a2/...)}}
{{redirect-distinguish|Recurring fraction|Repeating decimal}}

A '''simple''' or '''regular continued fraction''' is a [[continued fraction]] with numerators all equal one, and denominators built from a sequence <math>\{a_i\}</math> of integer numbers. The sequence can be finite or infinite, resulting in a '''finite''' (or '''terminated''') continued fraction like
:<math>a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2 + \cfrac{1}{ \ddots + \cfrac{1}{a_n} }}}</math>
or an '''infinite''' continued fraction like
:<math>a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2 + \cfrac{1}{ \ddots }}}</math>

Typically, such a continued fraction is obtained through an [[iterative]] process of representing a number as the sum of its [[integer part]] and the [[multiplicative inverse|reciprocal]] of another number, then writing this other number as the sum of its integer part and another reciprocal, and so on. In the ''finite'' case, the iteration/[[recursion]] is stopped after finitely many steps by using an integer in lieu of another continued fraction. In contrast, an ''infinite'' continued fraction is an [[infinite expression (mathematics)|infinite expression]]. In either case, all integers in the sequence, other than the first, must be [[positive number|positive]]. The integers <math>a_i</math> are called the [[coefficient]]s or terms of the continued fraction.{{sfn|Pettofrezzo|Byrkit|1970|p=150}}

Simple continued fractions have a number of remarkable properties related to the [[Euclidean algorithm]] for integers or [[real number]]s. Every [[rational number]] {{sfrac|<math>p</math>|<math>q</math>}} has two closely related expressions as a finite continued fraction, whose coefficients {{mvar|a<sub>i</sub>}} can be determined by applying the Euclidean algorithm to <math>(p,q)</math>. The numerical value of an infinite continued fraction is [[irrational number|irrational]]; it is defined from its infinite sequence of integers as the [[limit (mathematics)|limit]] of a sequence of values for finite continued fractions. Each finite continued fraction of the sequence is obtained by using a finite [[Prefix (computer science)|prefix]] of the infinite continued fraction's defining sequence of integers. Moreover, every irrational number <math>\alpha</math> is the value of a ''unique'' infinite regular continued fraction, whose coefficients can be found using the non-terminating version of the Euclidean algorithm applied to the [[Commensurability (mathematics)|incommensurable]] values <math>\alpha</math> and 1. This way of expressing real numbers (rational and irrational) is called their ''continued fraction representation''.

==Motivation and notation==
Consider, for example, the [[rational number]] {{sfrac|415|93}}, which is around 4.4624. As a first [[approximation]], start with 4, which is the [[Floor and ceiling functions|integer part]]; {{nowrap|1={{sfrac|415|93}} = 4 + {{sfrac|43|93}}}}. The fractional part is the [[Multiplicative inverse|reciprocal]] of {{sfrac|93|43}} which is about 2.1628. Use the integer part, 2, as an approximation for the reciprocal to obtain a second approximation of {{nowrap |1=4 + {{sfrac|1|2}} = 4.5. Now, {{sfrac|93|43}} = 2 + {{sfrac|7|43}}}};
the remaining fractional part, {{sfrac|7|43}}, is the reciprocal of {{sfrac|43|7}}, and {{sfrac|43|7}} is around 6.1429. Use 6 as an approximation for this to obtain {{nowrap|2 + {{sfrac|1|6}}}} as an approximation for {{sfrac|93|43}} and {{nowrap|4 + {{sfrac|1|2 + {{sfrac|1|6}}}}}}, about 4.4615, as the third approximation. Further, {{nowrap|1={{sfrac|43|7}} = 6 + {{sfrac|1|7}}}}. Finally, the fractional part, {{sfrac|1|7}}, is the reciprocal of 7, so its approximation in this scheme, 7, is exact ({{nowrap|1={{sfrac|7|1}} = 7 + {{sfrac|0|1}}}}) and produces the exact expression <math display=block> 4 + \cfrac1{2 + \cfrac1{6 + \cfrac17}} </math> for {{sfrac|415|93}}.

That expression is called the continued fraction representation of {{sfrac|415|93}}. This can be represented by the abbreviated notation {{sfrac|415|93}} = [4; 2, 6, 7]. (It is customary to replace only the ''first'' comma by a semicolon to indicate that the preceding number is the whole part.) Some older textbooks use all commas in the {{math|(''n'' + 1)}}-tuple, for example, [4, 2, 6, 7].{{sfn|Long|1972|p=173}}{{sfn|Pettofrezzo|Byrkit|1970|p=152}}

If the starting number is rational, then this process exactly parallels the [[Euclidean algorithm]] applied to the numerator and denominator of the number. In particular, it must terminate and produce a finite continued fraction representation of the number. The sequence of integers that occur in this representation is the sequence of successive quotients computed by the Euclidean algorithm. If the starting number is [[Irrational number|irrational]], then the process continues indefinitely. This produces a sequence of approximations, all of which are rational numbers, and these converge to the starting number as a limit. This is the (infinite) continued fraction representation of the number. Examples of continued fraction representations of irrational numbers are:
* {{math|1={{sqrt|19}} = [4;2,1,3,1,2,8,2,1,3,1,2,8,...]}} {{OEIS|A010124}}. The pattern repeats indefinitely with a period of 6.
* {{math|1=[[e (mathematical constant)|''e'']] = [2;1,2,1,1,4,1,1,6,1,1,8,...]}} {{OEIS|A003417}}. The pattern repeats indefinitely with a period of 3 except that 2 is added to one of the terms in each cycle.
* {{math|1=[[Pi|π]] = [3;7,15,1,292,1,1,1,2,1,3,1,...]}} {{OEIS|A001203}}. No pattern has ever been found in this representation.
* {{math|1=[[golden ratio|Φ]] = [1;1,1,1,1,1,1,1,1,1,1,1,...]}} {{OEIS|A000012}}. The [[golden ratio]], the irrational number that is the "most difficult" to approximate rationally {{crossreference|(see [[#A property of the golden ratio φ|{{nowrap|§{{tsp}}A}} property of the golden ratio φ]] below)}}.
* {{math|1=[[Euler–Mascheroni constant|γ]] = [0;1,1,2,1,2,1,4,3,13,5,1,...]}} {{OEIS|A002852}}. The [[Euler–Mascheroni constant]], which is expected but not known to be irrational, and whose continued fraction has no apparent pattern.

Continued fractions are, in some ways, more "mathematically natural" representations of a [[real number]] than other representations such as [[decimal representation]]s, and they have several desirable properties:
* The continued fraction representation for a real number is finite if and only if it is a rational number. In contrast, the decimal representation of a rational number may be finite, for example {{nowrap|1={{sfrac|137|1600}} = 0.085625}}, or infinite with a repeating cycle, for example {{nowrap|1={{sfrac|4|27}} = 0.148148148148...}}
* Every rational number has an essentially unique simple continued fraction representation. Each rational can be represented in exactly two ways, since {{math|1=[''a''<sub>0</sub>;''a''<sub>1</sub>,... ''a''<sub>''n''−1</sub>,''a''<sub>''n''</sub>] = [''a''<sub>0</sub>;''a''<sub>1</sub>,... ''a''<sub>''n''−1</sub>,(''a''<sub>''n''</sub>−1),1]}}. Usually the first, shorter one is chosen as the [[canonical form|canonical representation]].
* The simple continued fraction representation of an irrational number is unique. (However, additional representations are possible when using ''generalized'' continued fractions; see below.) 
* The real numbers whose continued fraction eventually repeats are precisely the [[quadratic irrational]]s.{{sfn|Weisstein|2022}} For example, the repeating continued fraction {{nowrap|[1;1,1,1,...]}} is the [[golden ratio]], and the repeating continued fraction {{nowrap|[1;2,2,2,...]}} is the [[square root of 2]]. In contrast, the decimal representations of quadratic irrationals are apparently [[normal number|random]]. The square roots of all (positive) integers that are not perfect squares are quadratic irrationals, and hence are unique periodic continued fractions.
* The successive approximations generated in finding the continued fraction representation of a number, that is, by truncating the continued fraction representation, are in a certain sense (described below) the "best possible".

== Formulation ==
A continued fraction in canonical form is an expression of the form

:<math>a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2 + \cfrac{1}{a_3 + \vphantom{\cfrac11} {_\ddots}}}}</math>

where ''a<sub>i</sub>'' are integer numbers, called the ''coefficients'' or ''terms'' of the continued fraction.{{sfn|Pettofrezzo|Byrkit|1970|p=150}}

When the expression contains finitely many terms, it is called a ''finite'' continued fraction.
When the expression contains infinitely many terms, it is called an ''infinite'' continued fraction.{{sfn|Collins|2001}}
When the terms eventually repeat from some point onwards, the continued fraction is called ''[[periodic continued fraction|periodic]]''.{{sfn|Weisstein|2022}}

Thus, all of the following illustrate valid finite simple continued fractions:

{| class="wikitable"
|+Examples of finite simple continued fractions
!Formula
!Numeric
!Remarks
|-
|<math>\ a_0</math>
|<math>\ 2</math>
|All integers are a [[degenerate case]]
|-
|<math>\ a_0 + \cfrac{1}{a_1}</math>
|<math>\ 2 + \cfrac{1}{3}</math>
|Simplest possible fractional form
|-
|<math>\ a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2}}</math>
|<math>\ -3 + \cfrac{1}{2 + \cfrac{1}{18}}</math>
|First integer may be negative
|-
|<math>\ a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2 + \cfrac{1}{a_3}}}</math>
|<math>\ \cfrac{1}{15 + \cfrac{1}{1 + \cfrac{1}{102}}}</math>
|First integer may be zero
|}
For simple continued fractions of the form

: <math>r=a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2 + \cfrac{1}{a_3 + \vphantom{\cfrac11} {_\ddots}}}}</math>

the <math>a_n</math> term can be calculated using the following recursive formula:

: <math>a_n= \left \lfloor \frac{N_n}{N_{n+1}} \right \rfloor</math>

where <math>N_{n+1}=N_{n-1} \bmod N_n </math> and <math>\begin{cases} N_0=r, \\ N_1=1, \end{cases}
</math>

from which it can be understood that the <math>a_n</math> sequence stops if <math>N_{n+1} = 0</math>.

===Notations===

Consider a continued fraction expressed as

:<math>x = a_0 + \cfrac{1}{a_1 + \cfrac{1}{a_2 + \cfrac{1}{a_3 + \cfrac{1}{a_4}}}}</math>

Because such a continued fraction expression may take a significant amount of vertical space, a number of methods have been tried to shrink it.

[[Gottfried Wilhelm Leibniz|Gottfried Leibniz]] sometimes used the notation<ref>{{cite journal |last=Cajori |first=Florian |author-link=Florian Cajori |year=1925 |title=Leibniz, the Master-Builder of Mathematical Notations |journal=Isis |volume=7 |number=3 |pages=412–429 |doi=10.1086/358328 |doi-access=free }}</ref>

<!-- very hacky workaround; mediawiki latex doesn't have good support for vertical alignment commands -->
:<math>
\begin{align}x = a_0 + \dfrac{1}{a_1} {{}\atop+} \\[28mu]\ \end{align} \!
\begin{align} \dfrac{1}{a_2} {{}\atop+} \\[2mu] \ \end{align} \!
\begin{align}\dfrac{1}{a_3} {{}\atop+}\end{align} \!
\begin{align}\\[2mu] \dfrac{1}{a_4},\end{align}
</math>

and later the same idea was taken even further with the nested fraction bars drawn aligned, for example by [[Alfred Pringsheim]] as

:<math>
x = a_0
+ {{}\atop{\big|\!}}\! \frac{1}{\,a_1} \!{{\!\big|}\atop{}}
+ {{}\atop{\big|\!}}\! \frac{1}{\,a_2} \!{{\!\big|}\atop{}}
+ {{}\atop{\big|\!}}\! \frac{1}{\,a_3} \!{{\!\big|}\atop{}}
+ {{}\atop{\big|\!}}\! \frac{1}{\,a_4} \!{{\!\big|}\atop{}},</math>
<!-- simpler coded alternative which doesn't look as good (though both are somewhat hacky):
:<math>
x = a_0 + \frac{\ \ 1 \ \mid}{\mid\,a_1}
+ \frac{\ \ 1\,\mid}{\mid\,a_2}
+ \frac{\ \ 1\,\mid}{\mid\,a_3}
+ \frac{\ \ 1\,\mid}{\mid\,a_4},</math>
-->
or in more common related notations as<ref>{{cite book |last=Swanson |first=Ellen |year=1999 |orig-year=1971 |title=Mathematics into Type |edition=Updated  |publisher=American Mathematical Society |others=Updated by O'Sean, Arlene; Schleyer, Antoinette |url=https://www.ams.org/publications/authors/mit-2.pdf |at=2.4.1c "Continued fractions", {{pgs|18}}}}</ref>

:<math>
x = a_0 +
{1 \over a_1 +}\,
{1 \over a_2 +}\,
{1 \over a_3 +}\,
{1 \over a_4}
</math>

or

:<math>
x = a_0 +
{1 \over a_1} {{}\atop+}
{1 \over a_2} {{}\atop+}
{1 \over a_3} {{}\atop+}
{1 \over a_4}.
</math>

[[Carl Friedrich Gauss]] used a notation reminiscent of [[Summation#Capital-sigma notation|summation notation]],

:<math>x = a_0 + \underset{i=1}{\overset{4}{\mathrm K}} ~ \frac{1}{a_i}, </math>

or in cases where the numerator is always 1, eliminated the fraction bars altogether, writing a list-style

:<math>x = [a_0; a_1, a_2, a_3, a_4]. </math>

Sometimes list-style notation uses angle brackets instead,

:<math>x = \left\langle a_0; a_1, a_2, a_3, a_4 \right\rangle.</math>

The semicolon in the square and angle bracket notations is sometimes replaced by a comma.{{sfn|Long|1972|p=173}}{{sfn|Pettofrezzo|Byrkit|1970|p=152}}

One may also define ''infinite simple continued fractions'' as [[limit of a sequence|limits]]:

:<math>[a_0; a_1, a_2, a_3, \,\ldots\, ] = \lim_{n \to \infty}\, [a_0; a_1, a_2, \,\ldots, a_n]. </math>

This limit exists for any choice of <math>a_0</math> and positive integers <math>a_1,a_2,\ldots</math>.{{sfn|Long|1972|p=183}}{{sfn|Pettofrezzo|Byrkit|1970|p=158}}

==Calculating continued fraction representations==
Consider a real number {{tmath|r}}.
Let <math>i=\lfloor r \rfloor </math> and let {{tmath|1=f = r - i}}.
When {{tmath|f\neq 0}}, the continued fraction representation of <math>r</math> is
{{tmath|[i;a_1,a_2,\ldots]}}, where <math>[a_1;a_2,\ldots]</math> is the continued fraction representation of {{tmath|1/f}}. When {{tmath|r\geq 0}}, then <math>i</math> is the [[integer part]] of <math>r</math>, and <math>f</math> is the [[fractional part]] of {{tmath|r}}.

In order to calculate a continued fraction representation of a number <math>r</math>, write down the [[Floor function|floor]] of <math>r</math>. Subtract this value from <math>r</math>. If the difference is 0, stop; otherwise find the [[multiplicative inverse|reciprocal]] of the difference and repeat. The procedure will halt if and only if <math>r</math> is rational. This process can be efficiently implemented using the [[Euclidean algorithm]] when the number is rational.

The table below shows an implementation of this procedure for the number {{tmath|1= 3.245 = 649/200 }}:

:{| class="wikitable"
|-
!Step
!Real<br/>Number
!Integer<br/>part
!Fractional<br/>part
!Simplified
!Reciprocal<br/>of {{mvar|f}}
|-
!1
|<math>r = \frac{649}{200}</math>
|<math>i = 3</math>
|<math>f = \frac{649}{200} - 3 </math>
|<math>= \frac{49}{200}</math>
|<math>\frac{1}{f} = \frac{200}{49} </math>
|-
!2
|<math>r = \frac{200}{49} </math>
|<math>i = 4</math>
|<math>f = \frac{200}{49} - 4 </math>
|<math> = \frac{4}{49} </math>
|<math> \frac{1}{f} = \frac{49}{4} </math>
|-
!3
|<math>r = \frac{49}{4} </math>
|<math>i = 12</math>
|<math>f = \frac{49}{4} - 12 </math>
|<math>= \frac{1}{4} </math>
|<math>\frac{1}{f} = \frac{4}{1} </math>
|-
!4
|<math>r = 4</math>
|<math>i = 4 </math>
|<math>f = 4 - 4 </math>
|<math>= 0 </math>
|colspan="2"|'''STOP'''
|}

The continued fraction for {{tmath|3.245}} is thus <math>[3; 4,12,4],</math> or, expanded:

<math display=block>
\frac{649}{200} = 3 + \cfrac{1}{4 + \cfrac{1}{12 + \cfrac{1}{4}}}.
</math>

==Reciprocals==
The continued fraction representations of a positive rational number and its [[multiplicative inverse|reciprocal]] are identical except for a shift one place left or right depending on whether the number is less than or greater than one respectively. In other words, the numbers represented by
<math>[a_0;a_1,a_2,\ldots,a_n]</math> and <math>[0;a_0,a_1,\ldots,a_n]</math> are reciprocals.

For instance if <math>a</math> is an integer and <math>x < 1</math> then

:<math>x=0 + \frac{1}{a + \frac{1}{b}}</math> and <math>\frac{1}{x} = a + \frac{1}{b}</math>.
If <math>x>1</math> then
:<math>x = a + \frac{1}{b}</math> and <math>\frac{1}{x} = 0 + \frac{1}{a + \frac{1}{b}}</math>.

The last number that generates the remainder of the continued fraction is the same for both <math>x</math> and its reciprocal.

For example,
:<math>2.25 = \frac{9}{4} = [2;4]</math> and <math>\frac{1}{2.25} = \frac{4}{9} = [0;2,4]</math>.

==Finite continued fractions==<!-- This section is linked from [[Continued fraction]] -->
Every finite continued fraction represents a [[rational number]], and every rational number can be represented in precisely two different ways as a finite continued fraction, with the conditions that the first coefficient is an integer and the other coefficients are positive integers. These two representations agree except in their final terms. In the longer representation the final term in the continued fraction is 1; the shorter representation drops the final 1, but increases the new final term by 1. The final element in the short representation is therefore always greater than 1, if present. In symbols:

:{{math|[''a''{{sub|0}}; ''a''{{sub|1}}, ''a''{{sub|2}}, ..., ''a''{{sub|''n'' − 1}}, ''a''{{sub|''n''}}, 1] {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ''a''{{sub|2}}, ..., ''a''{{sub|''n'' − 1}}, ''a''{{sub|''n''}} + 1]}}.
:{{math|[''a''{{sub|0}}; 1] {{=}} [''a''{{sub|0}} + 1]}}.

== Infinite continued fractions and convergents {{anchor|Convergents}} ==
[[File:Golden ration convergents.svg|thumb|Convergents approaching the [[golden ratio]]]]
Every infinite continued fraction is [[irrational number|irrational]], and every irrational number can be represented in precisely one way as an infinite continued fraction.

An infinite continued fraction representation for an irrational number is useful because its initial segments provide rational approximations to the number. These rational numbers are called the '''convergents''' of the continued fraction.{{sfn|Long|1972|p=177}}{{sfn|Pettofrezzo|Byrkit|1970|pp=162–163}} The larger a term is in the continued fraction, the closer the corresponding convergent is to the irrational number being approximated. Numbers like π have occasional large terms in their continued fraction, which makes them easy to approximate with rational numbers. Other numbers like ''e'' have only small terms early in their continued fraction, which makes them more difficult to approximate rationally. The [[golden ratio]] Φ has terms equal to 1 everywhere—the smallest values possible—which makes Φ the most difficult number to approximate rationally. In this sense, therefore, it is the "most irrational" of all irrational numbers. Even-numbered convergents are smaller than the original number, while odd-numbered ones are larger.

For a continued fraction {{math|[''a''{{sub|0}}; ''a''{{sub|1}}, ''a''{{sub|2}}, ...]}}, the first four convergents (numbered 0 through 3) are

:<math>
\frac{a_0}{1},\,
\frac{a_1a_0 + 1}{a_1},\,
\frac{a_2(a_1a_0 + 1) + a_0}{a_2a_1 + 1},\,
\frac{a_3\bigl(a_2(a_1a_0 + 1) + a_0\bigr) + (a_1a_0 + 1)}{a_3(a_2a_1 + 1) + a_1}.
</math>

The numerator of the third convergent is formed by multiplying the numerator of the second convergent by the third coefficient, and adding the numerator of the first convergent. The denominators are formed similarly. Therefore, each convergent can be expressed explicitly in terms of the continued fraction as the ratio of certain [[multivariate polynomial]]s called ''[[Continuant (mathematics)|continuants]]''.

If successive convergents are found, with numerators {{mvar|h}}{{sub|1}}, {{mvar|h}}{{sub|2}}, ... and denominators {{mvar|k}}{{sub|1}}, {{mvar|k}}{{sub|2}}, ... then the relevant recursive relation is that of [[Gaussian brackets]]:

:<math>\begin{align}
h_n &= a_nh_{n-1} + h_{n-2}, \\[3mu]
k_n &= a_nk_{n-1} + k_{n-2}.
\end{align}</math>

The successive convergents are given by the formula

:<math>\frac{h_n}{k_n} = \frac{a_nh_{n-1} + h_{n-2}}{a_nk_{n-1} + k_{n-2}}.</math>

Thus to incorporate a new term into a rational approximation, only the two previous convergents are necessary. The initial "convergents" (required for the first two terms) are <sup>0</sup>⁄<sub>1</sub> and <sup>1</sup>⁄<sub>0</sub>. For example, here are the convergents for [0;1,5,2,2].

:{| class="wikitable"
|- align="right"
! {{mvar|n}}
| −2|| −1|| 0 || 1 || 2 || 3 || 4
|- align="right"
! {{math|''a''{{sub|''n''}}}}
| &nbsp; || &nbsp; || 0 || 1 || 5 || 2 || 2
|- align="right"
! {{math|''h''{{sub|''n''}}}}
| 0 || 1 || 0 || 1 || 5 || 11 || 27
|- align="right"
! {{math|''k''{{sub|''n''}}}}
| 1 || 0 || 1 || 1 || 6 || 13 || 32
|}

When using the [[Methods of computing square roots#Heron's method|Babylonian method]] to generate successive approximations to the square root of an integer, if one starts with the lowest integer as first approximant, the rationals generated all appear in the list of convergents for the continued fraction. Specifically, the approximants will appear on the convergents list in positions 0, 1, 3, 7, 15, ... , {{math|2<sup>''k''</sup>−1}}, ... For example, the continued fraction expansion for [[square root of 3|<math>\sqrt3</math>]] is {{math|[1;&thinsp;1,&thinsp;2,&thinsp;1,&thinsp;2,&thinsp;1,&thinsp;2,&thinsp;1,&thinsp;2,&thinsp;...]}}. Comparing the convergents with the approximants derived from the Babylonian method:

:{| class="wikitable"
|- align="right"
! {{mvar|n}}
| −2|| −1|| '''0''' || '''1''' || 2 || '''3''' || 4 || 5 || 6 || '''7'''
|- align="right"
! {{math|''a''{{sub|''n''}}}}
| &nbsp; || &nbsp; || 1 || 1 || 2 || 1 || 2 || 1 || 2 || 1
|- align="right"
! {{math|''h''{{sub|''n''}}}}
| 0 || 1 || '''1''' || '''2''' || 5 || '''7''' || 19 || 26 || 71 || '''97'''
|- align="right"
! {{math|''k''{{sub|''n''}}}}
| 1 || 0 || '''1''' || '''1''' || 3 || '''4''' || 11 || 15 || 41 || '''56'''
|}

:{{math|''x''{{sub|0}} {{=}} 1 {{=}} {{sfrac|1|1}}}}
:{{math|''x''{{sub|1}} {{=}} {{sfrac|1|2}}(1 + {{sfrac|3|1}}) {{=}} {{sfrac|2|1}} {{=}} 2}}
:{{math|''x''{{sub|2}} {{=}} {{sfrac|1|2}}(2 + {{sfrac|3|2}}) {{=}} {{sfrac|7|4}}}}
:{{math|''x''{{sub|3}} {{=}} {{sfrac|1|2}}({{sfrac|7|4}} + {{sfrac|3|{{sfrac|7|4}}}}) {{=}} {{sfrac|97|56}}}}

===Properties===
The [[Baire space (set theory)|Baire space]] is a topological space on infinite sequences of natural numbers. The infinite continued fraction provides a [[homeomorphism]] from the Baire space to the space of irrational real numbers (with the subspace topology inherited from the [[Euclidean topology|usual topology]] on the reals). The infinite continued fraction also provides a map between the [[quadratic irrational]]s and the [[dyadic rational]]s, and from other irrationals to the set of infinite strings of binary numbers (i.e. the [[Cantor set]]); this map is called the [[Minkowski question-mark function]]. The mapping has interesting self-similar [[fractal]] properties; these are given by the [[modular group]], which is the subgroup of [[Möbius transformation]]s having integer values in the transform. Roughly speaking, continued fraction convergents can be taken to be Möbius transformations acting on the (hyperbolic) [[upper half-plane]]; this is what leads to the fractal self-symmetry.

The limit probability distribution of the coefficients in the continued fraction expansion of a random variable uniformly distributed in (0, 1) is the [[Gauss–Kuzmin distribution]].

===Some useful theorems===
If <math>\ a_0\ ,</math> <math>a_1\ ,</math> <math>a_2\ ,</math> <math>\ \ldots\ </math> is an infinite sequence of positive integers, define the sequences <math>\ h_n\ </math> and <math>\ k_n\ </math> recursively:
{| border="0" cellpadding="5" cellspacing="10" align="none"
|-
|<math> h_{n} = a_n\ h_{n-1} + h_{n-2}\ ,</math>
|
|
|<math> h_{-1} = 1\ ,</math>
|
|<math> h_{-2} = 0\ ;</math>
|-
|<math> k_{n}= a_n\ k_{n-1} + k_{n-2}\ ,</math>
|
|
|<math> k_{-1} = 0\ ,</math>
|
|<math> k_{-2} = 1 ~.</math>
|}

<blockquote>'''Theorem 1.''' For any positive real number <math>\ x\ </math>

:<math> \left[\ a_0;\ a_1,\ \dots, a_{n-1}, x\ \right]=\frac{x\ h_{n-1} + h_{n-2}}{\ x\ k_{n-1} + k_{n-2}\ }, \quad \left[\ a_0;\ a_1,\ \dots, a_{n-1} + x\ \right]=\frac{h_{n-1} + xh_{n-2}}{\ k_{n-1} + x k_{n-2}\ }</math>
</blockquote>

<blockquote>'''Theorem 2.''' The convergents of <math>\ [\ a_0\ ;</math> <math>a_1\ ,</math> <math>a_2\ ,</math> <math>\ldots\ ]\ </math> are given by

:<math>\left[\ a_0;\ a_1,\ \dots, a_n\ \right] = \frac{h_n}{\ k_n\ } ~.</math>
or in matrix form,<math display="block">\begin{bmatrix}
h_n & h_{n-1} \\
k_n & k_{n-1}
\end{bmatrix} = \begin{bmatrix}
a_0 & 1 \\
1 & 0
\end{bmatrix} \cdots 
\begin{bmatrix}
a_n & 1 \\
1 & 0
\end{bmatrix}</math>

'''Theorem 3.''' If the <math>\ n</math>th convergent to a continued fraction is <math>\ \frac{ h_n}{ k_n }\ ,</math> then
:<math> k_n\ h_{n-1} - k_{n-1}\ h_n = (-1)^n\ ,</math>
or equivalently
:<math> \frac{ h_n }{\ k_n\ } - \frac{ h_{n-1} }{\ k_{n-1}\ } = \frac{ (-1)^{n+1} }{\ k_{n-1}\ k_n\ } ~.</math>
</blockquote>

'''Corollary 1:''' Each convergent is in its lowest terms (for if <math>\ h_n\ </math> and <math>\ k_n\ </math> had a nontrivial common divisor it would divide <math>\ k_n\ h_{n-1} - k_{n-1}\ h_n\ ,</math> which is impossible).

'''Corollary 2:''' The difference between successive convergents is a fraction whose numerator is unity:

:<math> \frac{h_n}{k_n} - \frac{ h_{n-1} }{ k_{n-1} } = \frac{\ h_n\ k_{n-1} - k_n\ h_{n-1}\ }{\ k_n\ k_{n-1}\ } = \frac{ (-1)^{n+1} }{\ k_n\ k_{n-1}\ } ~.</math>

'''Corollary 3:''' The continued fraction is equivalent to a series of alternating terms:

:<math>a_0 + \sum_{n=0}^\infty \frac{ (-1)^n }{\ k_{n}\ k_{n+1}\ } ~.</math>

'''Corollary 4:''' The matrix
:<math>\begin{bmatrix}
h_n & h_{n-1} \\
k_n & k_{n-1}
\end{bmatrix} = \begin{bmatrix}
a_0 & 1 \\
1 & 0
\end{bmatrix} \cdots 
\begin{bmatrix}
a_n & 1 \\
1 & 0
\end{bmatrix}</math>
has [[determinant]] <math>(-1)^{n+1}</math>, and thus belongs to the group of
<math>\ 2\times 2\ </math> [[unimodular matrix|unimodular matrices]] <math>\ \mathrm{GL}(2,\mathbb{Z}) ~.</math>

'''Corollary 5:''' The matrix<math display="block">\begin{bmatrix}
h_n & h_{n-2} \\
k_n & k_{n-2}
\end{bmatrix} = \begin{bmatrix}
h_{n-1} & h_{n-2} \\
k_{n-1} & k_{n-2}
\end{bmatrix}
\begin{bmatrix}
a_{n} & 0 \\
1 & 1
\end{bmatrix}</math>
has determinant <math>(-1)^na_n</math>, or equivalently,<math display="block"> \frac{ h_n }{\ k_n\ } - \frac{ h_{n-2} }{\ k_{n-2}\ } = \frac{ (-1)^{n} }{\ k_{n-2 }\ k_n\ }a_n</math>meaning that the odd terms monotonically decrease, while the even terms monotonically increase.

'''Corollary 6:''' The denominator sequence <math>k_0, k_1, k_2, \dots</math> satisfies the recurrence relation <math>k_{-1} = 0, k_0 = 1, k_n = k_{n-1}a_n + k_{n-2}</math>, and grows at least as fast as the [[Fibonacci sequence]], which itself grows like <math>O(\phi^n)</math> where <math>\phi= 1.618\dots</math> is the [[golden ratio]].

<blockquote>'''Theorem 4.''' Each (<math>\ s</math>th) convergent is nearer to a subsequent (<math>\ n</math>th) convergent than any preceding (<math>\ r</math>th) convergent is. In symbols, if the <math>\ n</math>th convergent is taken to be <math>\ \left[\ a_0;\ a_1,\ \ldots,\ a_n\ \right] = x_n\ ,</math> then
:<math> \left|\ x_r - x_n\ \right| > \left|\ x_s - x_n\ \right| </math>
for all <math>\ r < s < n ~.</math> </blockquote>

'''Corollary 1:''' The even convergents (before the <math>\ n</math>th) continually increase, but are always less than <math>\ x_n ~.</math>

'''Corollary 2:''' The odd convergents (before the <math>\ n</math>th) continually decrease, but are always greater than <math>\ x_n ~.</math>

<blockquote>'''Theorem 5.'''

:<math>\frac{1}{\ k_n\ (k_{n+1} + k_n)\ } < \left|\ x - \frac{ h_n }{\ k_n\ }\ \right| < \frac{1}{\ k_n\ k_{n+1}\ } ~.</math>
</blockquote>

'''Corollary 1:''' A convergent is nearer to the limit of the continued fraction than any fraction whose denominator is less than that of the convergent.

'''Corollary 2:''' A convergent obtained by terminating the continued fraction just before a large term is a close approximation to the limit of the continued fraction.<blockquote>'''Theorem 6:''' Consider the set of all open intervals with end-points <math>[0;a_1, \dots, a_n], [0;a_1, \dots, a_n+1]</math>. Denote it as <math>\mathcal C</math>. Any open subset of <math>[0, 1] \setminus \Q</math> is a disjoint union of sets from <math>\mathcal C</math>.</blockquote>'''Corollary:''' The infinite continued fraction provides a homeomorphism from the Baire space to <math>[0, 1] \setminus \Q</math>.

==Semiconvergents==<!-- This section is linked from [[Complete quotient]] -->

If

:<math> \frac{h_{n-1}}{k_{n-1}},\frac{h_n}{k_n} </math>

are consecutive convergents, then any fractions of the form

: <math> \frac{h_{n-1} + mh_n}{k_{n-1} + mk_n},</math>

where <math>m</math> is an integer such that <math>0\leq m\leq a_{n+1}</math>, are called ''semiconvergents'', ''secondary convergents'', or ''intermediate fractions''. The <math>(m+1)</math>-st semiconvergent equals the [[Mediant (mathematics)|mediant]] of the <math>m</math>-th one and the convergent <math>\tfrac{h_n}{k_n}</math>. Sometimes the term is taken to mean that being a semiconvergent excludes the possibility of being a convergent (i.e., <math>0<m<a_{n+1}</math>), rather than that a convergent is a kind of semiconvergent.

It follows that semiconvergents represent a [[monotonic sequence]] of fractions between the convergents <math>\tfrac{h_{n-1}}{k_{n-1}}</math> (corresponding to <math>m=0</math>) and <math>\tfrac{h_{n+1}}{k_{n+1}}</math> (corresponding to <math>m=a_{n+1}</math>). The consecutive semiconvergents <math>\tfrac{a}{b}</math> and <math>\tfrac{c}{d}</math> satisfy the property <math>ad - bc = \pm 1</math>.

If a [[Diophantine approximation|rational approximation]] <math>\tfrac{p}{q}</math> to a real number <math>x</math> is such that the value <math>\left|x-\tfrac{p}{q}\right|</math> is smaller than that of any approximation with a smaller denominator, then <math>\tfrac{p}{q}</math> is a semiconvergent of the continued fraction expansion of <math>x</math>. The converse is not true, however.

==Best rational approximations==
{{See also|Diophantine approximation|Padé approximant}}

One can choose to define a ''best rational approximation'' to a real number {{mvar|x}} as a rational number {{sfrac|{{mvar|n}}|{{mvar|d}}}}, {{math|''d'' > 0}}, that is closer to {{mvar|x}} than any approximation with a smaller or equal denominator. The simple continued fraction for {{mvar|x}} can be used to generate ''all'' of the best rational approximations for {{mvar|x}} by applying these three rules:

#Truncate the continued fraction, and reduce its last term by a chosen amount (possibly zero).
#The reduced term cannot have less than half its original value.
#If the final term is even, half its value is admissible only if the corresponding semiconvergent is better than the previous convergent. (See below.)

For example, 0.84375 has continued fraction [0;1,5,2,2]. Here are all of its best rational approximations.

:{| class="wikitable"
|- align="center"
! Continued fraction
|  [0;1]  ||  [0;1,3]  ||  [0;1,4]  ||  [0;1,5]  ||  [0;1,5,2]  ||  [0;1,5,2,1]  ||  [0;1,5,2,2] 
|- align="center"
! Rational approximation
| 1 || {{sfrac|3|4}} || {{sfrac|4|5}} || {{sfrac|5|6}} || {{sfrac|11|13}} || {{sfrac|16|19}} || {{sfrac|27|32}}
|- align="center"
! Decimal equivalent
| 1 || 0.75 || 0.8 || ~0.83333 || ~0.84615 || ~0.84211 || 0.84375
|- align="center"
! Error
| +18.519% || −11.111% || −5.1852% || −1.2346% || +0.28490% || −0.19493% || 0%
|}
{{Diophantine_approximation_graph.svg}}

The strictly monotonic increase in the denominators as additional terms are included permits an algorithm to impose a limit, either on size of denominator or closeness of approximation.

The "half rule" mentioned above requires that when {{mvar|a}}{{sub|{{mvar|k}}}} is even, the halved term {{mvar|a}}{{sub|{{mvar|k}}}}/2 is admissible if and only if {{math|{{!}}''x'' − [''a''{{sub|0}} ; ''a''{{sub|1}}, ..., ''a''{{sub|''k'' − 1}}]{{!}} > {{!}}''x'' − [''a''{{sub|0}} ; ''a''{{sub|1}}, ..., ''a''{{sub|''k'' − 1}}, ''a''{{sub|''k''}}/2]{{!}}}} {{sfn|Thill|2008}} This is equivalent {{sfn|Thill|2008}} to: {{harvtxt|Shoemake|1995}}.

:{{math|[''a''{{sub|''k''}}; ''a''{{sub|''k'' − 1}}, ..., ''a''{{sub|1}}] > [''a''{{sub|''k''}}; ''a''{{sub|''k'' + 1}}, ...]}}.

The convergents to {{mvar|x}} are "best approximations" in a much stronger sense than the one defined above. Namely, {{mvar|n}}/{{mvar|d}} is a convergent for {{mvar|x}} if and only if {{math|{{!}}''dx'' − ''n''{{!}}}} has the smallest value among the analogous expressions for all rational approximations {{mvar|m}}/{{mvar|c}} with {{math|''c'' ≤ ''d''}}; that is, we have {{math|{{!}}''dx'' − ''n''{{!}} < {{!}}''cx'' − ''m''{{!}}}} so long as {{math|''c'' < ''d''}}. (Note also that {{math|{{!}}''d<sub>k</sub>x'' − ''n<sub>k</sub>''{{!}} → 0}} as {{math|''k'' → ∞}}.)

=== Best rational within an interval ===
A rational that falls within the interval {{open-open|''x'', ''y''}}, for {{math|0 < {{mvar|x}} < {{mvar|y}}}}, can be found with the continued fractions for {{mvar|x}} and {{mvar|y}}. When both {{mvar|x}} and {{mvar|y}} are irrational and
:{{math|''x'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ''a''{{sub|2}}, ..., ''a''{{sub|''k'' − 1}}, ''a''{{sub|''k''}}, ''a''{{sub|''k'' + 1}}, ...]}}
:{{math|''y'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ''a''{{sub|2}}, ..., ''a''{{sub|''k'' − 1}}, ''b''{{sub|''k''}}, ''b''{{sub|''k'' + 1}}, ...]}}
where {{mvar|x}} and {{mvar|y}} have identical continued fraction expansions up through {{math|''a''<sub>''k''−1</sub>}}, a rational that falls within the interval {{open-open|''x'', ''y''}} is given by the finite continued fraction,
:{{math|''z''(''x'',''y'') {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ''a''{{sub|2}}, ..., ''a''{{sub|''k'' − 1}}, min(''a''{{sub|''k''}}, ''b''{{sub|''k''}}) + 1]}}
This rational will be best in the sense that no other rational in {{open-open|''x'', ''y''}} will have a smaller numerator or a smaller denominator.<ref>{{cite web
 | last = Gosper | first = R. W. | author-link = Bill Gosper
 | title = Appendix 2: Continued Fraction Arithmetic
 | url = https://perl.plover.com/yak/cftalk/INFO/gosper.ps
 | year = 1977}} See "simplest intervening rational", pp. 29–31.</ref><ref>{{cite journal
 | last = Murakami | first = Hiroshi
 | date = February 2015
 | doi = 10.1145/2733693.2733711
 | issue = 3/4
 | journal = ACM Communications in Computer Algebra
 | pages = 134–136
 | title = Calculation of rational numbers in an interval whose denominator is the smallest by using FP interval arithmetic
 | volume = 48}}</ref>

If {{mvar|x}} is rational, it will have ''two'' continued fraction representations that are ''finite'', {{math|''x''<sub>1</sub>}} and {{math|''x''<sub>2</sub>}}, and similarly a rational&nbsp;{{mvar|y}} will have two representations, {{math|''y''<sub>1</sub>}} and {{math|''y''<sub>2</sub>}}. The coefficients beyond the last in any of these representations should be interpreted as {{math|+∞}}; and the best rational will be one of {{math|''z''(''x''<sub>1</sub>, ''y''<sub>1</sub>)}}, {{math|''z''(''x''<sub>1</sub>, ''y''<sub>2</sub>)}}, {{math|''z''(''x''<sub>2</sub>, ''y''<sub>1</sub>)}}, or {{math|''z''(''x''<sub>2</sub>, ''y''<sub>2</sub>)}}.

For example, the decimal representation 3.1416 could be rounded from any number in the interval {{closed-open|3.14155, 3.14165}}. The continued fraction representations of 3.14155 and 3.14165 are
:{{math|3.14155 {{=}} [3; 7, 15, 2, 7, 1, 4, 1, 1] {{=}} [3; 7, 15, 2, 7, 1, 4, 2]}}
:{{math|3.14165 {{=}} [3; 7, 16, 1, 3, 4, 2, 3, 1] {{=}} [3; 7, 16, 1, 3, 4, 2, 4]}}
and the best rational between these two is
:{{math|[3; 7, 16] {{=}} {{sfrac|355|113}} {{=}} 3.1415929....}}
Thus, {{sfrac|355|113}} is the best rational number corresponding to the rounded decimal number 3.1416, in the sense that no other rational number that would be rounded to 3.1416 will have a smaller numerator or a smaller denominator.

=== Interval for a convergent ===
A rational number, which can be expressed as finite continued fraction in two ways,
:{{math|''z'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ..., ''a''{{sub|''k'' − 1}}, ''a''{{sub|''k''}}, 1] {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ..., ''a''{{sub|''k'' − 1}}, ''a''{{sub|''k''}} + 1] {{=}} {{sfrac| ''p''{{sub|''k''}}|''q''{{sub|''k''}}}}}}
will be one of the convergents for the continued fraction expansion of a number, if and only if the number is strictly between (see [https://math.stackexchange.com/a/4438961 this proof])
:{{math|''x'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ..., ''a''{{sub|''k'' − 1}}, ''a''{{sub|''k''}}, 2] {{=}} {{sfrac| 2''p''{{sub|''k''}} - ''p''{{sub|''k-1''}}|2''q''{{sub|''k''}} - ''q''{{sub|''k-1''}}}}}} and
:{{math|''y'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ..., ''a''{{sub|''k'' − 1}}, ''a''{{sub|''k''}} + 2] {{=}} {{sfrac| ''p''{{sub|''k''}} + ''p''{{sub|''k-1''}}|''q''{{sub|''k''}} + ''q''{{sub|''k-1''}}}}}}
The numbers {{mvar|x}} and {{mvar|y}} are formed by incrementing the last coefficient in the two representations for {{mvar|z}}. It is the case that {{math|''x'' < ''y''}} when {{mvar|k}} is even, and {{math|''x'' > ''y''}} when {{mvar|k}} is odd.

For example, the number {{sfrac|355|113}} has the continued fraction representations
:{{sfrac|355|113}} = [3; 7, 15, 1] = [3; 7, 16]
and thus {{sfrac|355|113}} is a convergent of any number strictly between
:{| cellpadding="2" cellspacing="0"
| align="right" | {{math|[3; 7, 15, 2]}} ||{{=}}|| {{math|{{sfrac|688|219}} ≈ 3.1415525}}
|-
| align="right" | {{math|[3; 7, 17]}} ||{{=}}|| {{math|{{sfrac|377|120}} ≈ 3.1416667}}
|}

=== Legendre's theorem on continued fractions ===
{{see also|Dirichlet's approximation theorem}}
In his ''Essai sur la théorie des nombres'' (1798), [[Adrien-Marie Legendre]] derives a necessary and sufficient condition for a rational number to be a convergent of the continued fraction of a given real number.<ref>{{cite book|last=Legendre|first=Adrien-Marie|author-link=Adrien-Marie Legendre|title=Essai sur la théorie des nombres|date=1798|publisher=Duprat|location=Paris|publication-date=1798|pages=27–29|language=fr}}</ref> A consequence of this criterion, often called '''Legendre's theorem''' within the study of continued fractions, is as follows:<ref>{{cite journal|last1=Barbolosi|first1=Dominique|last2=Jager|first2=Hendrik|date=1994|title=On a theorem of Legendre in the theory of continued fractions|url=https://www.jstor.org/stable/26273940|journal=[[Journal de Théorie des Nombres de Bordeaux]]|volume=6|issue=1|pages=81–94|doi=10.5802/jtnb.106 |jstor=26273940 |via=JSTOR}}</ref>

'''Theorem'''. If ''α'' is a real number and ''p'', ''q'' are positive integers such that <math>\left|\alpha - \frac{p}{q}\right| < \frac{1}{2q^2}</math>, then ''p''/''q'' is a convergent of the continued fraction of ''α''.
{{collapse top|title = Proof}}
'''Proof'''. We follow the proof given in ''[[An Introduction to the Theory of Numbers]]'' by [[G. H. Hardy]] and [[E. M. Wright]].<ref>{{cite book|last1=Hardy|first1=G. H.|author-link=G. H. Hardy|last2=Wright|first2=E. M.|author-link2=E. M. Wright|title=An Introduction to the Theory of Numbers|title-link=An Introduction to the Theory of Numbers|publisher=[[Oxford University Press]]|year=1938|isbn=|location=London|publication-date=1938|pages=140–141, 153|language=en}}</ref>

Suppose ''α'', ''p'', ''q'' are such that <math>\left|\alpha - \frac{p}{q}\right| < \frac{1}{2q^2}</math>, and assume that ''α'' > ''p''/''q''. Then we may write <math>\alpha - \frac{p}{q} = \frac{\theta}{q^2}</math>, where 0 < ''θ'' < 1/2. We write ''p''/''q'' as a finite continued fraction [''a''<sub>0</sub>; ''a''<sub>1</sub>, ..., ''a<sub>n</sub>''], where due to the fact that each rational number has two distinct representations as finite continued fractions differing in length by one (namely, one where ''a<sub>n</sub>'' = 1 and one where ''a<sub>n</sub>'' ≠ 1), we may choose ''n'' to be even. (In the case where ''α'' < ''p''/''q'', we would choose ''n'' to be odd.)

Let ''p''<sub>0</sub>/''q''<sub>0</sub>, ..., ''p<sub>n</sub>''/''q<sub>n</sub>'' = ''p''/''q'' be the convergents of this continued fraction expansion. Set <math>\omega := \frac{1}{\theta} - \frac{q_{n-1}}{q_n}</math>, so that <math>\theta = \frac{q_n}{q_{n-1} + \omega q_n}</math> and thus,<math display="block">\alpha = \frac{p}{q} + \frac{\theta}{q^2} = \frac{p_n}{q_n} + \frac{1}{q_n (q_{n-1} + \omega q_n)} = \frac{(p_n q_{n-1} + 1) + \omega p_n q_n}{q_n (q_{n-1} + \omega q_n)} = \frac{p_{n-1} q_n + \omega p_n q_n}{q_n (q_{n-1} + \omega q_n)} = \frac{p_{n-1} + \omega p_n}{q_{n-1} + \omega q_n}, </math>where we have used the fact that ''p<sub>n</sub>''<sub>-1</sub> ''q<sub>n</sub>'' - ''p<sub>n</sub>'' ''q<sub>n</sub>''<sub>-1</sub> = (-1)''<sup>n</sup>'' and that ''n'' is even.

Now, this equation implies that ''α'' = [''a''<sub>0</sub>; ''a''<sub>1</sub>, ..., ''a<sub>n</sub>'', ''ω'']. Since the fact that 0 < ''θ'' < 1/2 implies that ''ω'' > 1, we conclude that the continued fraction expansion of ''α'' must be [''a''<sub>0</sub>; ''a''<sub>1</sub>, ..., ''a<sub>n</sub>'', ''b''<sub>0</sub>, ''b''<sub>1</sub>, ...], where [''b''<sub>0</sub>; ''b''<sub>1</sub>, ...] is the continued fraction expansion of ''ω'', and therefore that ''p<sub>n</sub>''/''q<sub>n</sub>'' = ''p''/''q'' is a convergent of the continued fraction of ''α''.
{{collapse bottom}}

This theorem forms the basis for [[Wiener's attack]], a polynomial-time exploit of the [[RSA (cryptosystem)|RSA cryptographic protocol]] that can occur for an injudicious choice of public and private keys (specifically, this attack succeeds if the prime factors of the public key ''n'' = ''pq'' satisfy ''p'' < ''q'' < 2''p'' and the private key ''d'' is less than (1/3)''n''<sup>1/4</sup>).<ref>{{cite journal|last=Wiener|first=Michael J.|date=1990|title=Cryptanalysis of short RSA secret exponents|url=https://ieeexplore.ieee.org/document/54902|journal=[[IEEE Transactions on Information Theory]]|volume=36|issue=3|pages=553–558|doi=10.1109/18.54902 |via=IEEE}}</ref>

==Comparison==
Consider {{math|''x'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ...]}} and {{math|''y'' {{=}} [''b''{{sub|0}}; ''b''{{sub|1}}, ...]}}. If {{mvar|k}} is the smallest index for which {{math|''a''{{sub|''k''}}}} is unequal to {{math|''b''{{sub|''k''}}}} then {{math|''x'' < ''y''}} if {{math|(−1){{sup|''k''}}(''a''{{sub|''k''}} − ''b''{{sub|''k''}}) < 0}} and {{math|''y'' < ''x''}} otherwise.

If there is no such {{mvar|k}}, but one expansion is shorter than the other, say {{math|''x'' {{=}} [''a''{{sub|0}}; ''a''{{sub|1}}, ..., ''a''{{sub|''n''}}]}} and {{math|''y'' {{=}} [''b''{{sub|0}}; ''b''{{sub|1}}, ..., ''b''{{sub|''n''}}, ''b''{{sub|''n'' + 1}}, ...]}} with {{math|''a''{{sub|''i''}} {{=}} ''b''{{sub|''i''}}}} for {{math|0 ≤ ''i'' ≤ ''n''}}, then {{math|''x'' < ''y''}} if {{mvar|n}} is even and {{math|''y'' < ''x''}} if {{mvar|n}} is odd.

==Continued fraction expansion of {{pi}} and its convergents==
To calculate the convergents of [[pi|{{pi}}]] we may set {{math|''a''{{sub|0}} {{=}} ⌊{{pi}}⌋ {{=}} 3}}, define {{math|''u''{{sub|1}} {{=}} {{sfrac|1|{{pi}} − 3}} ≈ 7.0625}} and {{math|''a''{{sub|1}} {{=}} ⌊''u''{{sub|1}}⌋ {{=}} 7}}, {{math|''u''{{sub|2}} {{=}} {{sfrac|1|''u''{{sub|1}} − 7}} ≈ 15.9966}} and {{math|''a''{{sub|2}} {{=}} ⌊''u''{{sub|2}}⌋ {{=}} 15}}, {{math|''u''{{sub|3}} {{=}} {{sfrac|1|''u''{{sub|2}} − 15}} ≈ 1.0034}}. Continuing like this, one can determine the infinite continued fraction of {{pi}} as
:[3;7,15,1,292,1,1,...] {{OEIS|A001203}}.
The fourth convergent of {{pi}} is [3;7,15,1] = {{sfrac|355|113}} = 3.14159292035..., sometimes called [[Milü]], which is fairly close to the true value of {{pi}}.

Let us suppose that the quotients found are, as above, [3;7,15,1]. The following is a rule by which we can write down at once the convergent fractions which result from these quotients without developing the continued fraction.

The first quotient, supposed divided by unity, will give the first fraction, which will be too small, namely, {{sfrac|3|1}}. Then, multiplying the numerator and denominator of this fraction by the second quotient and adding unity to the numerator, we shall have the second fraction, {{sfrac|22|7}}, which will be too large. Multiplying in like manner the numerator and denominator of this fraction by the third quotient, and adding to the numerator the numerator of the preceding fraction, and to the denominator the denominator of the preceding fraction, we shall have the third fraction, which will be too small. Thus, the third quotient being 15, we have for our numerator {{nowrap|(22 × 15 {{=}} 330) + 3 {{=}} 333}}, and for our denominator, {{nowrap|(7 × 15 {{=}} 105) + 1 {{=}} 106}}. The third convergent, therefore, is {{sfrac|333|106}}. We proceed in the same manner for the fourth convergent. The fourth quotient being 1, we say 333 times 1 is 333, and this plus 22, the numerator of the fraction preceding, is 355; similarly, 106 times 1 is 106, and this plus 7 is 113.
In this manner, by employing the four quotients [3;7,15,1], we obtain the four fractions:
:{{sfrac|3|1}}, {{sfrac|22|7}}, {{sfrac|333|106}}, {{sfrac|355|113}}, ....

[[File:Continued fraction expansion of Pi .png|thumb|The following Maple code will generate continued fraction expansions of pi]]

To sum up, the pattern is
<small><math> \text{Numerator}_i = \text{Numerator}_{(i-1)} \cdot \text{Quotient}_i + \text{Numerator}_{(i-2)} </math>
<math> \text{Denominator}_i = \text{Denominator}_{(i-1)} \cdot \text{Quotient}_i + \text{Denominator}_{(i-2)} </math></small>

These convergents are alternately smaller and larger than the true value of {{pi}}, and approach nearer and nearer to {{pi}}. The difference between a given convergent and {{pi}} is less than the reciprocal of the product of the denominators of that convergent and the next convergent. For example, the fraction {{sfrac|22|7}} is greater than {{pi}}, but {{sfrac|22|7}} − {{pi}} is less than {{sfrac|1|7 × 106}}&nbsp;=&nbsp;{{sfrac|1|742}} (in fact, {{sfrac|22|7}} − {{pi}} is just more than {{sfrac|1|791}} = {{sfrac|1|7 × 113}}).

The demonstration of the foregoing properties is deduced from the fact that if we seek the difference between one of the convergent fractions and the next adjacent to it we shall obtain a fraction of which the numerator is always unity and the denominator the product of the two denominators. Thus the difference between {{sfrac|22|7}} and {{sfrac|3|1}} is {{sfrac|1|7}}, in excess; between {{sfrac|333|106}} and {{sfrac|22|7}}, {{sfrac|1|742}}, in deficit; between {{sfrac|355|113}} and {{sfrac|333|106}}, {{sfrac|1|11978}}, in excess; and so on. The result being, that by employing this series of differences we can express in another and very simple manner the fractions with which we are here concerned, by means of a second series of fractions of which the numerators are all unity and the denominators successively be the product of every two adjacent denominators. Instead of the fractions written above, we have thus the series:

:{{sfrac|3|1}} + {{sfrac|1|1 × 7}} − {{sfrac|1|7 × 106}} + {{sfrac|1|106 × 113}} − ...

The first term, as we see, is the first fraction; the first and second together give the second fraction, {{sfrac|22|7}}; the first, the second and the third give the third fraction {{sfrac|333|106}}, and so on with the rest; the result being that the series entire is equivalent to the original value.

==Non-simple continued fraction==
{{main|Continued fraction (non-simple)}}

A non-simple continued fraction is an expression of the form

:<math>x = b_0 + \cfrac{a_1}{b_1 + \cfrac{a_2}{b_2 + \cfrac{a_3}{b_3 + \cfrac{a_4}{b_4 + \ddots\,}}}}</math>

where the ''a''<sub>''n''</sub> (''n'' &gt; 0) are the partial numerators, the ''b''<sub>''n''</sub> are the partial denominators, and the leading term ''b''<sub>0</sub> is called the ''integer'' part of the continued fraction.

To illustrate the use of non-simple continued fractions, consider the following example. The sequence of partial denominators of the simple continued fraction of {{pi}} does not show any obvious pattern:

:<math>\pi=[3;7,15,1,292,1,1,1,2,1,3,1,\ldots]</math>

or

:<math>\pi=3+\cfrac{1}{7+\cfrac{1}{15+\cfrac{1}{1+\cfrac{1}{292+\cfrac{1}{1+\cfrac{1}{1+\cfrac{1}{1+\cfrac{1}{2+\cfrac{1}{1+\cfrac{1}{3+\cfrac{1}{1+\ddots}}}}}}}}}}}</math>

However, several non-simple continued fractions for {{pi}} have a perfectly regular structure, such as:

:<math>
\pi=\cfrac{4}{1+\cfrac{1^2}{2+\cfrac{3^2}{2+\cfrac{5^2}{2+\cfrac{7^2}{2+\cfrac{9^2}{2+\ddots}}}}}}
=\cfrac{4}{1+\cfrac{1^2}{3+\cfrac{2^2}{5+\cfrac{3^2}{7+\cfrac{4^2}{9+\ddots}}}}}
=3+\cfrac{1^2}{6+\cfrac{3^2}{6+\cfrac{5^2}{6+\cfrac{7^2}{6+\cfrac{9^2}{6+\ddots}}}}}
</math>

:<math>\displaystyle \pi=2+\cfrac{2}{1+\cfrac{1}{1/2+\cfrac{1}{1/3+\cfrac{1}{1/4+\ddots}}}}=2+\cfrac{2}{1+\cfrac{1\cdot2}{1+\cfrac{2\cdot3}{1+\cfrac{3\cdot4}{1+\ddots}}}}</math>

:<math> \displaystyle \pi=2+\cfrac{4}{3+\cfrac{1\cdot3}{4+\cfrac{3\cdot5}{4+\cfrac{5\cdot7}{4+\ddots}}}}</math>

The first two of these are special cases of the [[Inverse trigonometric functions#Variant: Continued fractions for arctangent|arctangent]] function with {{pi}} = 4 arctan (1) and the fourth and fifth one can be derived using the [[Wallis product]].{{sfn|Bunder|Tonien|2017}}{{sfn|Scheinerman|Pickett|Coleman|2008}}

:<math>
\pi=3+\cfrac{1}{6+\cfrac{1^3+2^3}{6\cdot1^2+1^2\cfrac{1^3+2^3+3^3+4^3}{6\cdot2^2+2^2\cfrac{1^3+2^3 +3^3+4^3+5^3+6^3}{6\cdot3^2+3^2\cfrac{1^3+2^3+3^3+4^3+5^3+6^3+7^3+8^3}{6\cdot4^2+\ddots}}}}}
</math>

The continued fraction of <math>\pi</math> above consisting of cubes uses the Nilakantha series and an exploit from Leonhard Euler.{{sfn|Foster|2015}}

==Other continued fraction expansions==

===Periodic continued fractions===
{{main|Periodic continued fraction}}
The numbers with periodic continued fraction expansion are precisely the [[quadratic irrational|irrational solutions]] of [[quadratic equation]]s with rational coefficients; rational solutions have finite continued fraction expansions as previously stated. The simplest examples are the [[golden ratio]] φ = [1;1,1,1,1,1,...] and {{sqrt|2}} = [1;2,2,2,2,...], while {{sqrt|14}} = [3;1,2,1,6,1,2,1,6...] and {{sqrt|42}} = [6;2,12,2,12,2,12...]. All irrational square roots of integers have a special form for the period; a symmetrical string, like the empty string (for {{sqrt|2}}) or 1,2,1 (for {{sqrt|14}}), followed by the double of the leading integer.

===A property of the golden ratio φ===
Because the continued fraction expansion for [[golden ratio|φ]] doesn't use any integers greater than 1, φ is one of the most "difficult" real numbers to approximate with rational numbers. [[Hurwitz's theorem (number theory)|Hurwitz's theorem]] {{sfn|Hardy|Wright|2008|loc=Theorem 193}} states that any irrational number {{mvar|k}} can be approximated by infinitely many rational {{sfrac|''m''|''n''}} with

:<math>\left| k - {m \over n}\right| < {1 \over n^2 \sqrt 5}.</math>

While virtually all real numbers {{mvar|k}} will eventually have infinitely many convergents {{sfrac|''m''|''n''}} whose distance from {{mvar|k}} is significantly smaller than this limit, the convergents for φ (i.e., the numbers {{sfrac|5|3}}, {{sfrac|8|5}}, {{sfrac|13|8}}, {{sfrac|21|13}}, etc.) consistently "toe the boundary", keeping a distance of almost exactly <math>{\scriptstyle{1 \over n^2 \sqrt 5}}</math> away from φ, thus never producing an approximation nearly as impressive as, for example, [[Milü|{{sfrac|355|113}}]] for [[pi|{{pi}}]]. It can also be shown that every real number of the form {{sfrac|''a'' + ''b''φ|''c'' + ''d''φ}}, where {{mvar|a}}, {{mvar|b}}, {{mvar|c}}, and {{mvar|d}} are integers such that {{math|1=''a'' ''d'' − ''b'' ''c'' = ±1}}, shares this property with the golden ratio φ; and that all other real numbers can be more closely approximated.

===Regular patterns in continued fractions===
While there is no discernible pattern in the simple continued fraction expansion of {{pi}}, there is one for {{math|''e''}}, the [[e (mathematical constant)|base of the natural logarithm]]:

:<math>e = e^1 = [2; 1, 2, 1, 1, 4, 1, 1, 6, 1, 1, 8, 1, 1, 10, 1, 1, 12, 1, 1, \dots],</math>

which is a special case of this general expression for positive integer {{mvar|n}}:

:<math>e^{1/n} = [1; n-1, 1, 1, 3n-1, 1, 1, 5n-1, 1, 1, 7n-1, 1, 1, \dots] \,\!.</math>

Another, more complex pattern appears in this continued fraction expansion for positive odd {{mvar|n}}:

:<math>e^{2/n} = \left[1; \frac{n-1}{2}, 6n, \frac{5n-1}{2}, 1, 1, \frac{7n-1}{2}, 18n, \frac{11n-1}{2}, 1, 1, \frac{13n-1}{2}, 30n, \frac{17n-1}{2}, 1, 1, \dots \right] \,\!,</math>

with a special case for {{math|1=''n'' = 1}}:

:<math>e^2 = [7; 2, 1, 1, 3, 18, 5, 1, 1, 6, 30, 8, 1, 1, 9, 42, 11, 1, 1, 12, 54, 14, 1, 1 \dots, 3k, 12k+6, 3k+2, 1, 1 \dots] \,\!.</math>

Other continued fractions of this sort are

:<math>\tanh(1/n) = [0; n, 3n, 5n, 7n, 9n, 11n, 13n, 15n, 17n, 19n, \dots] </math>

where {{mvar|n}} is a positive integer; also, for integer {{mvar|n}}:

:<math>\tan(1/n) = [0; n-1, 1, 3n-2, 1, 5n-2, 1, 7n-2, 1, 9n-2, 1, \dots]\,\!,</math>

with a special case for {{math|1=''n'' = 1}}:

:<math>\tan(1) = [1; 1, 1, 3, 1, 5, 1, 7, 1, 9, 1, 11, 1, 13, 1, 15, 1, 17, 1, 19, 1, \dots]\,\!.</math>

If {{math|''I''<sub>''n''</sub>(''x'')}} is the modified, or hyperbolic, [[Bessel function]] of the first kind, we may define a function on the rationals {{sfrac|''p''|''q''}} by

:<math>S(p/q) = \frac{I_{p/q}(2/q)}{I_{1+p/q}(2/q)},</math>

which is defined for all rational numbers, with {{mvar|p}} and {{mvar|q}} in lowest terms. Then for all nonnegative rationals, we have

:<math>S(p/q) = [p+q; p+2q, p+3q, p+4q, \dots],</math>

with similar formulas for negative rationals; in particular we have

:<math>S(0) = S(0/1) = [1; 2, 3, 4, 5, 6, 7, \dots].</math>

Many of the formulas can be proved using [[Gauss's continued fraction]].

===Typical continued fractions===
Most irrational numbers do not have any periodic or regular behavior in their continued fraction expansion. Nevertheless, for [[almost all]] numbers on the unit interval, they have the same limit behavior.

The arithmetic average diverges: <math>\lim_{n\to\infty}\frac 1n \sum_{k=1}^n a_k = +\infty</math>, and so the coefficients grow arbitrarily large: <math>\limsup_n a_n = +\infty</math>. In particular, this implies that almost all numbers are well-approximable, in the sense that<math display="block">\liminf_{n\to\infty}  \left|
x - \frac{p_n}{q_n}
\right| q_n^2 = 0</math>[[Aleksandr Khinchin|Khinchin]] proved that the [[geometric mean]] of {{math|''a''<sub>''i''</sub>}} tends to a constant (known as [[Khinchin's constant]]):<math display="block">\lim_{n \rightarrow \infty } \left( a_1 a_2 ... a_n \right) ^{1/n} = 
K_0 = 2.6854520010\dots</math>[[Paul Lévy (mathematician)|Paul Lévy]] proved that the {{mvar|n}}th root of the denominator of the {{mvar|n}}th convergent converges to [[Lévy's constant]]
<math display="block">\lim_{n \rightarrow \infty } q_n^{1/n} = 
e^{\pi^2/(12\ln2)} = 3.2758\ldots</math>[[Lochs' theorem]] states that the convergents converge exponentially at the rate of<math display="block">\lim_{n\to\infty}\frac 1n \ln\left|
x - \frac{p_n}{q_n}
\right| = -\frac{\pi^2}{6\ln 2} </math>

==Applications==
===Pell's equation===
Continued fractions play an essential role in the solution of [[Pell's equation]]. For example, for positive integers {{mvar|p}} and {{mvar|q}}, and non-square {{mvar|n}}, it is true that if {{math|1=''p''<sup>2</sup> − ''nq''<sup>2</sup> = ±1}}, then {{math|{{sfrac|''p''|''q''}}}} is a convergent of the regular continued fraction for {{sqrt|{{mvar|n}}}}.  The converse holds if the period of the regular continued fraction for {{sqrt|{{mvar|n}}}} is 1, and in general the period describes which convergents give solutions to Pell's equation.{{sfn|Niven|Zuckerman|Montgomery|1991}}

===Dynamical systems===
Continued fractions also play a role in the study of [[dynamical system]]s, where they tie together the [[Farey sequence|Farey fractions]] which are seen in the [[Mandelbrot set]] with [[Minkowski's question-mark function]] and the [[modular group]] Gamma.

The backwards [[shift operator]] for continued fractions is the map {{math|''h''(''x'') {{=}} 1/{{mvar|x}} − ⌊1/{{mvar|x}}⌋}} called the '''[[Gauss–Kuzmin–Wirsing operator|Gauss map]]''', which lops off digits of a continued fraction expansion: {{math|''h''([0; ''a''{{sub|1}}, ''a''{{sub|2}}, ''a''{{sub|3}}, ...]) {{=}} [0; ''a''{{sub|2}}, ''a''{{sub|3}}, ...]}}. The [[transfer operator]] of this map is called the [[Gauss–Kuzmin–Wirsing operator]]. The distribution of the digits in continued fractions is given by the zero'th [[eigenvector]] of this operator, and is called the [[Gauss–Kuzmin distribution]].

==History==
* 300 BCE ''[[Euclid's Elements]]'' contains an algorithm for the [[greatest common divisor]], whose [[Euclidean algorithm|modern version]] generates a continued fraction as the sequence of quotients of successive [[Euclidean division]]s that occur in it.
* 499 The ''[[Aryabhatiya]]'' contains the solution of indeterminate equations using continued fractions
* 1572 [[Rafael Bombelli]], ''L'Algebra Opera'' – method for the extraction of square roots which is related to continued fractions
* 1613 [[Pietro Cataldi]], ''Trattato del modo brevissimo di trovar la radice quadra delli numeri'' – first notation for continued fractions
:Cataldi represented a continued fraction as <math>a_0</math> & <math>\frac{n_1}{d_1 \cdot}</math> & <math>\frac{n_2}{d_2 \cdot}</math> & <math>\frac{n_3}{d_3 \cdot}</math> with the dots indicating where the following fractions went.
* 1695 [[John Wallis]], ''Opera Mathematica'' – introduction of the term "continued fraction"
* 1737 [[Leonhard Euler]], ''De fractionibus continuis dissertatio'' – Provided the first then-comprehensive account of the properties of continued fractions, and included the first proof that the number [[e (mathematical constant)|e]] is irrational.{{sfn|Sandifer|2006}}
* 1748 Euler, ''[[List of important publications in mathematics#Introductio in analysin infinitorum|Introductio in analysin infinitorum]]''. Vol. I, Chapter 18 – proved the equivalence of a certain form of continued fraction and a generalized [[infinite series]], proved that every rational number can be written as a finite continued fraction, and proved that the continued fraction of an irrational number is infinite.{{sfn|Euler|1748}}
* 1761 [[Johann Lambert]] – gave the first proof of the irrationality of [[Pi|{{pi}}]] using a continued fraction for [[Trigonometric functions|tan(x)]].
* 1768 [[Joseph-Louis Lagrange]] – provided the general solution to Pell's equation using continued fractions similar to Bombelli's
* 1770 Lagrange – proved that [[quadratic irrational number|quadratic irrationals]] expand to [[periodic continued fraction]]s.
* 1813 [[Carl Friedrich Gauss]], ''Werke'', Vol. 3, pp.&nbsp;134–138 – derived a very general [[Gauss's continued fraction|complex-valued continued fraction]] via a clever identity involving the [[hypergeometric function]]
* 1892 [[Henri Padé]] defined [[Padé approximant]]
* 1972 [[Bill Gosper]] – First exact algorithms for continued fraction arithmetic.

==See also==
* {{annotated link|Complete quotient}}
* {{annotated link|Methods of computing square roots#Continued fraction expansion|Computing continued fractions of square roots}}
* {{annotated link|Egyptian fraction}}
* {{annotated link|Engel expansion}}
* {{annotated link|Euler's continued fraction formula}}
* {{annotated link|Iterated binary operation}}
* {{annotated link|Klein polyhedron}}
* {{annotated link|Mathematical constants by continued fraction representation}}
* {{annotated link|Restricted partial quotients}}
* {{annotated link|Stern–Brocot tree}}

==Notes==
{{Reflist}}

==References==
*{{cite journal 
|last1= Bunder
|first1= Martin W.
|last2= Tonien
|first2= Joseph
|year=2017 
|title=Closed form expressions for two harmonic continued fractions 
|journal=The Mathematical Gazette
|volume=101
|issue=552
|pages= 439–448 
|url=https://ro.uow.edu.au/cgi/viewcontent.cgi?article=1781&context=eispapers1
|doi=10.1017/mag.2017.125
|s2cid= 125489697
}}

*{{cite journal 
|last1= Chen 
|first1= Chen-Fan 
|last2= Shieh 
|first2= Leang-San 
|year= 1969
|title= Continued fraction inversion by Routh's Algorithm 
|journal= IEEE Trans. Circuit Theory 
|volume= 16 
|number= 2 
|pages= 197–202 
|doi= 10.1109/TCT.1969.1082925 
}}

*{{cite journal 
|last= Collins 
|first= Darren C.
|date= 2001
|title= Continued Fractions 
|journal= MIT Undergraduate Journal of Mathematics 
|url= http://www-math.mit.edu/phase2/UJM/vol1/COLLIN~1.PDF 
|url-status= dead 
|archive-url= https://web.archive.org/web/20011120064343/http://www-math.mit.edu/phase2/UJM/vol1/COLLIN~1.PDF 
|archive-date= 2001-11-20
}}

*{{cite book
|last1= Cuyt 
|first1= A. 
|last2= Brevik Petersen 
|first2= V. 
|last3= Verdonk 
|first3= B. 
|last4= Waadeland 
|first4= H. 
|last5= Jones 
|first5= W. B. 
|year= 2008 
|title= Handbook of Continued fractions for Special functions 
|publisher= Springer Verlag 
|isbn= 978-1-4020-6948-2
}}

*{{cite web
|author=Encyclopædia Britannica
|title= Continued fraction – mathematics
|date= 2013
|url= https://www.britannica.com/science/continued-fraction
|access-date= 26 April 2022 |ref={{harvid|Encyclopaedia Britannica|2013}}
}}

*{{cite web
|last= Euler
|first= Leonhard
|date= 1748
|title=E101 – Introductio in analysin infinitorum, volume 1
|url= https://scholarlycommons.pacific.edu/euler-works/101/
|publisher= The Euler Archive
|access-date= 26 April 2022
}}

*{{Cite web
|last= Foster
|first= Tony
|date= 22 June 2015
|url= https://www.theoremoftheday.org/NumberTheory/EulerCF/TotDEulerCF.pdf |archive-url=https://web.archive.org/web/20131211085948/http://theoremoftheday.org/NumberTheory/EulerCF/TotDEulerCF.pdf |archive-date=2013-12-11 |url-status=live
|title= Theorem of the Day: Theorem no. 203
|publisher= Robin Whitty
|access-date= 26 April 2022
}}

*{{cite journal
|last= Gragg 
|first= William B. 
|year= 1974
|title= Matrix interpretations and applications of the continued fraction algorithm 
|journal= Rocky Mountain J. Math. 
|volume= 4 
|number= 2 
|doi= 10.1216/RMJ-1974-4-2-213 
|page= 213 
|s2cid= 121378061 
|doi-access= free 
}}

*{{Cite book
|last1= Hardy
|first1= Godfrey H.
|last2= Wright
|first2= Edward M.
|date= December 2008
|orig-date= 1979
|title= An Introduction to the Theory of Numbers
|publisher= Oxford University Press
|edition= 6
|isbn= 9780199219865
}}

*{{cite journal
|last= Heilermann 
|first= J. B. H. 
|year= 1846
|title= Ueber die Verwandlung von Reihen in Kettenbrüche 
|journal= Journal für die reine und angewandte Mathematik 
|volume= 33 
|pages= 174–188 
|url= http://www.digizeitschriften.de/dms/resolveppn/?PID=PPN243919689_0033%7Clog13
}}

*{{Cite book 
|last1= Jones 
|first1= William B. 
|last2= Thron 
|first2= W. J. 
|year= 1980 
|title= Continued Fractions: Analytic Theory and Applications. Encyclopedia of Mathematics and its Applications. 
|publisher= Addison-Wesley Publishing Company 
|location= Reading. Massachusetts 
|volume= 11 
|isbn= 0-201-13510-8 
|url-access= registration 
|url= https://archive.org/details/continuedfractio0000jone 
}}

*{{cite book 
|last= Khinchin 
|first= A. Ya. 
|author-link= Aleksandr Khinchin 
|year= 1964 
|title= Continued Fractions 
|url= https://archive.org/details/continuedfractio00khin_0 
|url-access= registration 
|orig-year= Originally published in Russian, 1935 
|publisher= [[University of Chicago Press]] 
|isbn= 0-486-69630-8 
}}

*{{cite book
|last= Long 
|first= Calvin T. 
|date= December 1972
|title= Elementary Introduction to Number Theory 
|edition= 2nd 
|publisher= [[D. C. Heath and Company]] 
|location= Lexington 
|isbn= 9780669627039 
}}

*{{cite journal
|last= Magnus 
|first= Arne 
|year= 1962
|title= Continued fractions associated with the Padé Table 
|journal= Math. Z. 
|volume= 78 
|pages= 361–374 
|doi= 10.1007/BF01195180 
|s2cid= 120535167 
}}

*{{cite book
|last1= Niven
|first1= Ivan
|last2= Zuckerman
|first2= Herbert S.
|last3= Montgomery
|first3= Hugh L.
|date= 1991
|title= An introduction to the theory of numbers
|url= https://archive.org/details/introductiontoth0000nive
|url-access= registration
|edition= Fifth
|location= New York
|publisher= [[John Wiley & Sons|Wiley]]
|isbn= 0-471-62546-9
}}

*{{cite book 
|last= Perron 
|first= Oskar 
|author-link= Oskar Perron 
|year= 1950
|title= Die Lehre von den Kettenbrüchen 
|publisher= Chelsea Publishing Company 
|place= New York, NY 
}}

*{{cite book
|last1= Pettofrezzo 
|first1= Anthony J. 
|last2= Byrkit 
|first2= Donald R. 
|date= December 1970
|title= Elements of Number Theory 
|publisher= [[Prentice Hall]] 
|location= Englewood Cliffs 
|isbn= 9780132683005 
}}

*{{cite journal 
|last = Rieger 
|first = Georg Johann 
|year = 1982 
|title = A new approach to the real numbers (motivated by continued fractions) 
|journal = Abhandlungen der Braunschweigischen Wissenschaftlichen Gesellschaft
|url=https://publikationsserver.tu-braunschweig.de/servlets/MCRFileNodeServlet/dbbs_derivate_00031201/Rieger_A_new_approach_to_the_real_numbers.pdf |archive-url=https://web.archive.org/web/20201210192036/https://publikationsserver.tu-braunschweig.de/servlets/MCRFileNodeServlet/dbbs_derivate_00031201/Rieger_A_new_approach_to_the_real_numbers.pdf |archive-date=2020-12-10 |url-status=live 
|volume = 33 
|pages = 205–217 
}}

*{{cite book 
|last1= Rockett 
|first1= Andrew M. 
|last2= Szüsz 
|first2= Peter
|year= 1992 
|title= Continued Fractions 
|url= https://archive.org/details/continuedfractio0000rock 
|url-access= registration 
|publisher= World Scientific Press 
|isbn= 981-02-1047-7 
}}

*{{cite book 
|last= Sandifer 
|first= C. Edward 
|year= 2006 
|title= How Euler Did It
|chapter= Chapter 32: Who proved ''e'' is irrational? 
|publisher= [[Mathematical Association of America]] 
|pages= 185–190 
|isbn= 978-0-88385-563-8 
|lccn= 2007927658
|url= http://eulerarchive.maa.org/hedi/HEDI-2006-02.pdf |archive-url=https://web.archive.org/web/20140319071616/http://eulerarchive.maa.org/hedi/HEDI-2006-02.pdf |archive-date=2014-03-19 |url-status=live
}}

*{{cite journal
|last1= Scheinerman
|first1= Ed
|last2= Pickett
|first2= Thomas J.
|last3= Coleman
|first3= Ann
|year=2008 
|title=Another Continued Fraction for π 
|journal= The American Mathematical Monthly
|volume=115
|issue=10
|pages= 930–933
|doi= 10.1080/00029890.2008.11920610
|jstor= 27642639
|s2cid= 11914017
}}

*{{cite book
|last= Shoemake
|first= Ken
|year= 1995
|editor-last= Paeth
|editor-first= Alan W.
|title= Graphics Gems V (IBM Version)
|chapter=I.4: Rational Approximation
|chapter-url= https://books.google.com/books?id=8CGj9_ZlFKoC&pg=PA25
|pages= 25–31
|publisher=Academic Press
|location=San Diego, California
|isbn= 0-12-543455-3
}}

*{{cite journal
|last= Siebeck 
|first= H. 
|year= 1846 
|title= Ueber periodische Kettenbrüche 
|journal= Journal für die reine und angewandte Mathematik 
|volume= 33 
|pages= 68–70 
|url= http://www.digizeitschriften.de/dms/resolveppn/?PID=PPN243919689_0033%7Clog5
}}

*{{cite journal 
|last= Thill 
|first=M.
|year=2008 
|title=A more precise rounding algorithm for rational numbers 
|journal=Computing 
|volume=82 
|issue=2–3
|pages=189–198 
|doi=10.1007/s00607-008-0006-7
|s2cid=45166490
}}

*{{cite web
|last=Thurston
|first=Ben
|title= Estimating square roots, generalized continued fraction expression for every square root
|publisher= The Ben Paul Thurston Blog
|date= 2012
|url= http://benpaulthurstonblog.blogspot.com/2012/05/estimating-square-roots.html
|access-date= 26 April 2022
}}

*{{cite book 
|last= Wall 
|first= Hubert Stanley
|author-link= Hubert Stanley Wall 
|year= 1948 
|title= Analytic Theory of Continued Fractions 
|publisher= D. Van Nostrand Company, Inc. 
|isbn= 0-8284-0207-8
}}

*{{cite web
|last= Weisstein
|first= Eric Wolfgang
|author-link= Eric W. Weisstein
|date= 2022
|title=Periodic Continued Fraction
|editor= MathWorld
|url=https://mathworld.wolfram.com/PeriodicContinuedFraction.html
|access-date= 26 April 2022
}}

==External links==
* {{springer|title=Continued fraction|id=p/c025540}}
*{{cite web
|last= Knott
|first= Ron
|date= 2018
|title= Continued fractions (An online Combined Continued Fraction Calculator is available)
|url= http://www.maths.surrey.ac.uk/hosted-sites/R.Knott/Fibonacci/cfINTRO.html
|access-date= 26 April 2022
}}

* Linas Vepstas [http://www.linas.org/math/chap-gap/chap-gap.html Continued Fractions and Gaps] (2004) reviews chaotic structures in continued fractions.
* [http://www.cut-the-knot.org/blue/ContinuedFractions.shtml Continued Fractions on the Stern-Brocot Tree] at [[cut-the-knot]]
* [http://www.math.sunysb.edu/~tony/whatsnew/column/antikytheraI-0400/kyth3.html The Antikythera Mechanism I: Gear ratios and continued fractions] {{Webarchive|url=https://web.archive.org/web/20090504105639/http://www.math.sunysb.edu/~tony/whatsnew/column/antikytheraI-0400/kyth3.html |date=2009-05-04 }}
* [http://wims.unice.fr/~wims/en_tool~number~contfrac.en.html Continued fraction calculator], WIMS.
* [https://web.archive.org/web/20030202011209/http://www.tweedledum.com/rwg/cfup.htm Continued Fraction Arithmetic] Gosper's first continued fractions paper, unpublished. Cached on the [[Internet Archive]]'s [[Internet Archive#Wayback Machine|Wayback Machine]]
* {{MathWorld |title=Continued Fraction |urlname=ContinuedFraction}}
* [http://demonstrations.wolfram.com/ContinuedFractions/ Continued Fractions] by [[Stephen Wolfram]] and [http://demonstrations.wolfram.com/ContinuedFractionApproximationsOfTheTangentFunction/ Continued Fraction Approximations of the Tangent Function] by Michael Trott, [[Wolfram Demonstrations Project]].
* {{OEIS el|1=A133593|2="Exact" continued fraction for pi}}
* [http://go.helms-net.de/math/tetdocs/FracIterAltGeom.htm A view into "fractional interpolation" of a continued fraction {{math|{1; 1, 1, 1, ...}}} ]
* [http://www.ams.org/publicoutreach/feature-column/fcarc-irrational3 Best rational approximation through continued fractions]
* [https://www.ms.uky.edu/~sohum/ma330/files/Continued%20Fractions.pdf CONTINUED FRACTIONS by C. D. Olds]
{{Wiktionary}}
{{Fractions and ratios}}

{{Authority control}}

[[Category:Continued fractions| ]]
[[Category:Mathematical analysis]]