Jump to content

Rouché's theorem: Difference between revisions

From Wikipedia, the free encyclopedia
Content deleted Content added
No edit summary
m Reverting possible vandalism by BazingaMan5000 to version by 159.149.2.125. Report False Positive? Thanks, ClueBot NG. (4362246) (Bot)
 
(46 intermediate revisions by 27 users not shown)
Line 1: Line 1:
{{Short description|Theorem about zeros of holomorphic functions}}
{{for|the theorem in linear algebra|Rouché–Capelli theorem}}
{{For|the theorem in linear algebra|Rouché–Capelli theorem}}
'''Rouché's theorem''', named after [[Eugène Rouché]], states that for any two [[complex number|complex]]-valued [[function (mathematics)|functions]] ''f'' and ''g'' [[Holomorphic function|holomorphic]] inside some region ''K'' with closed contour <math>\partial K</math>, if |''g''(''z'')|&nbsp;<&nbsp;|''f''(''z'')| on <math>\partial K</math>, then ''f'' and ''f''&nbsp;+&nbsp;''g'' have the same number of zeros inside ''K'', where each zero is counted as many times as its [[Multiplicity (mathematics)|multiplicity]]. This theorem assumes that the contour <math>\partial K</math> is simple, that is, without self-intersections. Rouché's theorem is an easy consequence of a stronger symmetric Rouché's theorem described below.
{{Complex analysis sidebar}}

'''Rouché's theorem''', named after [[Eugène Rouché]], states that for any two [[complex number|complex]]-valued [[function (mathematics)|functions]] {{mvar|f}} and {{math|''g''}} [[Holomorphic function|holomorphic]] inside some region <math>K</math> with closed contour <math>\partial K</math>, if {{math|{{!}}''g''(''z''){{!}} < {{!}}''f''(''z''){{!}}}} on <math>\partial K</math>, then {{math|''f''}} and {{math|''f'' + ''g''}} have the same number of zeros inside <math>K</math>, where each zero is counted as many times as its [[Multiplicity (mathematics)|multiplicity]]. This theorem assumes that the contour <math>\partial K</math> is simple, that is, without self-intersections. Rouché's theorem is an easy consequence of a stronger symmetric Rouché's theorem described below.
== Symmetric version ==

[[Theodor Estermann]] (1902–1991) proved in his book ''Complex Numbers and Functions'' the following statement: Let <math>K\subset G</math> be a bounded region with continuous boundary <math>\partial K</math>. Two holomorphic functions <math>f,\,g\in\mathcal H(G)</math> have the same number of roots (counting multiplicity) in <math>K</math>, if the strict inequality

:<math>|f(z)-g(z)|<|f(z)|+|g(z)| \qquad \left(z\in \partial K\right)</math>

holds on the boundary <math>\partial K</math>.

The original Rouché's theorem then follows by setting <math>f(z):=f(z)+g(z)</math> and <math>g(z):=f(z)</math>.


== Usage ==
== Usage ==
The theorem is usually used to simplify the problem of locating zeros, as follows. Given an analytic function, we write it as the sum of two parts, one of which is simpler and grows faster than (thus dominates) the other part. We can then locate the zeros by looking at only the dominating part. For example, the polynomial <math>z^5 + 3z^3 + 7</math> has exactly 5 zeros in the disk <math>|z| < 2</math> since <math>|3z^3 + 7| \le 31 < 32 = |z^5|</math> for every <math>|z| = 2</math>, and <math>z^5</math>, the dominating part, has five zeros in the disk.

The theorem is usually used to simplify the problem of locating zeros, as follows. Given an analytic function, we write it as the sum of two parts, one of which is simpler and grows faster than (thus dominates) the other part. We can then locate the zeros by looking at only the dominating part. For example, the polynomial <math>z^5 + 3z^3 + 7</math> has exactly 5 zeros in the disk <math>|z| < 2</math> since <math>|3z^3 + 7| < 32 = |z^5|</math> for every <math>|z| = 2</math>, and <math>z^5</math>, the dominating part, has five zeros in the disk.


== Geometric explanation ==
== Geometric explanation ==
[[Image:rouche-thm.png|thumb|300px|right|As ''z'' travels along a closed curve ''C'' (not shown in the picture), '''{{color|blue|''f''(''z'')}}''' and '''{{color|red|''h''(''z'')}}''' will trace out closed curves in the complex plane (shown in blue and red). So long as the curves never veer too far apart from each other (we require that '''{{color|blue|''f''(''z'')}}''' remains closer to '''{{color|red|''h''(''z'')}}''' than the origin at all times), then the curves will [[winding number|wind around the origin]] the same number of times. Then, by the [[argument principle]], '''{{color|blue|''f''(''z'')}}''' and '''{{color|red|''h''(''z'')}}''' have the same number of zeros inside ''C'' (not shown).]]
[[Image:rouche-thm.png|thumb|300px|right|Since the ''distance'' between the curves is ''small'', ''h''(''z'') does exactly one turn around just as ''f''(''z'') does.]]


It is possible to provide an informal explanation of Rouché's theorem.
It is possible to provide an informal explanation of Rouché's theorem.
Line 23: Line 14:
Let ''C'' be a closed, simple curve (i.e., not self-intersecting). Let ''h''(''z'') = ''f''(''z'') + ''g''(''z''). If ''f'' and ''g'' are both holomorphic on the interior of ''C'', then ''h'' must also be holomorphic on the interior of ''C''. Then, with the conditions imposed above, the Rouche's theorem in its original (and not symmetric) form says that
Let ''C'' be a closed, simple curve (i.e., not self-intersecting). Let ''h''(''z'') = ''f''(''z'') + ''g''(''z''). If ''f'' and ''g'' are both holomorphic on the interior of ''C'', then ''h'' must also be holomorphic on the interior of ''C''. Then, with the conditions imposed above, the Rouche's theorem in its original (and not symmetric) form says that


: If |''f''(''z'')| > |''h''(''z'')&nbsp;&minus;&nbsp;''f''(''z'')|, for every ''z'' in ''C,'' then ''f'' and ''h'' have the same number of zeros in the interior of ''C''.
{{block indent | em = 1.5 | text = If {{math|1={{!}}''f''(''z''){{!}} > {{!}}''h''(''z'') &minus; ''f''(''z''){{!}}}}, for every ''z'' in ''C,'' then ''f'' and ''h'' have the same number of zeros in the interior of ''C''.}}


Notice that the condition |''f''(''z'')| > |''h''(''z'')&nbsp;&minus;&nbsp;''f''(''z'')| means that for any ''z'', the distance from ''f''(''z'') to the origin is larger than the length of ''h''(''z'')&nbsp;&minus;&nbsp;''f''(''z''), which in the following picture means that for each point on the blue curve, the segment joining it to the origin is larger than the green segment associated with it. Informally we can say that the blue curve ''f''(''z'') is always closer to the red curve ''h''(''z'') than it is to the origin.
Notice that the condition |''f''(''z'')| > |''h''(''z'')&nbsp;&minus;&nbsp;''f''(''z'')| means that for any ''z'', the distance from ''f''(''z'') to the origin is larger than the length of ''h''(''z'')&nbsp;&minus;&nbsp;''f''(''z''), which in the following picture means that for each point on the blue curve, the segment joining it to the origin is larger than the green segment associated with it. Informally we can say that the blue curve ''f''(''z'') is always closer to the red curve ''h''(''z'') than it is to the origin.


The previous paragraph shows that ''h''(''z'') must wind around the origin exactly as many times as ''f''(''z''). The index of both curves around zero is therefore the same, so by the [[argument principle]], ''f''(''z'') and ''h''(''z'') must have the same number of zeros inside ''C''.
The previous paragraph shows that ''h''(''z'') must wind around the origin exactly as many times as ''f''(''z''). The index of both curves around zero is therefore the same, so by the [[argument principle]], {{math|''f''(''z'')}} and {{math|''h''(''z'')}} must have the same number of zeros inside {{mvar|C}}.


One popular, informal way to summarize this argument is as follows: If, say, [[Gérard de Nerval]] were to walk his lobster on a leash around and around a tree, such that the leash's length was always less than Nerval's distance to the tree, then the lobster would wind around the tree exactly as many times as Nerval.
One popular, informal way to summarize this argument is as follows: If a person were to walk a dog on a leash around and around a tree, such that the distance between the person and the tree is always greater than the length of the leash, then the person and the dog go around the tree the same number of times.


== Applications ==
== Applications ==
{{see also|Properties of polynomial_roots#Bounds on (complex) polynomial roots}}
{{See also|Properties of polynomial roots#Bounds on (complex) polynomial roots}}
Consider the polynomial <math>z^2 + 2az + b^2</math> (where <math>a > b > 0</math>). By the [[quadratic formula]] it has two zeros at <math>-a \pm \sqrt{a^2 - b^2}</math>. Rouché's theorem can be used to obtain more precise positions of them. Since
:<math>|z^2 + b^2| \le 2b^2 < 2a|z|</math> for every <math>|z| = b</math>,
Rouché's theorem says that the polynomial has exactly one zero inside the disk <math>|z| < b</math>. Since <math>a + \sqrt{a^2 - b^2}</math> is clearly outside the disk, we conclude that the zero is <math>a - \sqrt{a^2 - b^2}</math>. This sort of arguments can be useful in locating residues when one applies Cauchy's [[residue theorem]].


=== Bounding roots ===
Rouché's theorem can also be used to give a short proof of the [[fundamental theorem of algebra]]. Let
Consider the polynomial <math>z^2 + 2az + b^2</math> with <math>a > b > 0</math>. By the [[quadratic formula]] it has two zeros at <math>-a \pm \sqrt{a^2 - b^2}</math>. Rouché's theorem can be used to obtain some hint about their positions. Since
<math display="block">|z^2 + b^2| \le 2b^2 < 2a|z| \text{ for all } |z| = b,</math>


Rouché's theorem says that the polynomial has exactly one zero inside the disk <math>|z| < b</math>. Since <math>-a - \sqrt{a^2 - b^2}</math> is clearly outside the disk, we conclude that the zero is <math>-a + \sqrt{a^2 - b^2}</math>.
: <math>p(z) = a_0 + a_1z + a_2 z^2 + \cdots + a_n z^n, \quad a_n \ne 0\, </math>


In general, a polynomial <math>f(z) = a_n z^n + \cdots + a_0</math>. If <math>|a_k| r^k > \sum_{j\neq k}|a_j| r^j</math> for some <math>r > 0, k \in 0:n</math>, then by Rouche's theorem, the polynomial has exactly <math>k</math> roots inside <math>B(0, r)</math>.
and choose <math>R>0</math> so large that:


This sort of argument can be useful in locating residues when one applies Cauchy's [[residue theorem]].
:<math>|a_0 + a_1z + \cdots + a_{n-1}z^{n-1}| \le \sum_{j=0}^{n - 1} |a_j| R^{j} < |a_n|R^n = |a_n z^n|\text{ for }|z| = R.</math>


=== Fundamental theorem of algebra ===
Rouché's theorem can also be used to give a short proof of the [[fundamental theorem of algebra]]. Let
<math display="block">p(z) = a_0 + a_1z + a_2 z^2 + \cdots + a_n z^n, \quad a_n \ne 0</math>
and choose <math>R > 0</math> so large that:
<math display="block">|a_0 + a_1z + \cdots + a_{n-1} z^{n-1}| \le \sum_{j=0}^{n - 1} |a_j| R^j < |a_n| R^n = |a_n z^n| \text{ for } |z| = R.</math>
Since <math>a_n z^n</math> has <math>n</math> zeros inside the disk <math>|z| < R</math> (because <math>R>0</math>), it follows from Rouché's theorem that <math>p</math> also has the same number of zeros inside the disk.
Since <math>a_n z^n</math> has <math>n</math> zeros inside the disk <math>|z| < R</math> (because <math>R>0</math>), it follows from Rouché's theorem that <math>p</math> also has the same number of zeros inside the disk.


Line 51: Line 46:
Another use of Rouché's theorem is to prove the [[open mapping theorem (complex analysis)|open mapping theorem]] for analytic functions. We refer to the article for the proof.
Another use of Rouché's theorem is to prove the [[open mapping theorem (complex analysis)|open mapping theorem]] for analytic functions. We refer to the article for the proof.


== Symmetric version ==
==Proof of the symmetric form of Rouché's theorem==
A stronger version of Rouché's theorem was published by [[Theodor Estermann]] in 1962.<ref>{{cite book|last1=Estermann|first1=T.|title=Complex Numbers and Functions|date=1962|publisher=Athlone Press, Univ. of London|page=156}}</ref> It states: let <math>K\subset G</math> be a bounded region with continuous boundary <math>\partial K</math>. Two holomorphic functions <math>f,\,g\in\mathcal H(G)</math> have the same number of roots (counting multiplicity) in <math>K</math>, if the strict inequality
<math display="block">|f(z)-g(z)|<|f(z)|+|g(z)| \qquad \left(z\in \partial K\right)</math>
holds on the boundary <math>\partial K.</math>

The original version of Rouché's theorem then follows from this symmetric version applied to the functions <math>f+g,f</math> together with the trivial inequality <math>|f(z)+g(z)| \ge 0</math> (in fact this inequality is strict since <math>f(z)+g(z) = 0</math> for some <math>z\in\partial K</math> would imply <math>|g(z)| = |f(z)|</math>).

The statement can be understood intuitively as follows.
By considering <math>-g</math> in place of <math>g</math>, the condition can be rewritten as <math>|f(z) + g(z)|<|f(z)|+|g(z)|</math> for <math>z\in \partial K</math>.
Since <math>|f(z) + g(z)| \leq |f(z)|+|g(z)|</math> always holds by the triangle inequality, this is equivalent to saying that <math>|f(z) + g(z)| \neq |f(z)|+|g(z)|</math> on <math>\partial K</math>, which in turn means that for <math>z\in\partial K</math> the functions <math>f(z)</math> and <math>g(z)</math> are non-vanishing and <math>\arg{f(z)} \neq \arg{g(z)}</math>.

Intuitively, if the values of <math>f</math> and <math>g</math> never pass through the origin and never point in the same direction as <math>z</math> circles along <math>\partial K</math>, then <math>f(z)</math> and <math>g(z)</math> must wind around the origin the same number of times.

== Proof of the symmetric form of Rouché's theorem ==
Let <math>C\colon[0,1]\to\mathbb C</math> be a simple closed curve whose image is the boundary <math>\partial K</math>. The hypothesis implies that ''f'' has no roots on <math>\partial K</math>, hence by the [[argument principle]], the number ''N<sub>f</sub>''(''K'') of zeros of ''f'' in ''K'' is
Let <math>C\colon[0,1]\to\mathbb C</math> be a simple closed curve whose image is the boundary <math>\partial K</math>. The hypothesis implies that ''f'' has no roots on <math>\partial K</math>, hence by the [[argument principle]], the number ''N<sub>f</sub>''(''K'') of zeros of ''f'' in ''K'' is
:<math>\frac1{2\pi i}\oint_C\frac{f'(z)}{f(z)}\,dz=\frac1{2\pi i}\oint_{f\circ C}\frac{dz}z=\mathrm{Ind}_{f\circ C}(0),</math>
<math display="block">\frac1{2\pi i}\oint_C\frac{f'(z)}{f(z)}\,dz=\frac1{2\pi i}\oint_{f\circ C} \frac{dz}z =\mathrm{Ind}_{f\circ C}(0),</math>
i.e., the [[winding number]] of the closed curve <math>f\circ C</math> around the origin; similarly for ''g''. The hypothesis ensures that ''g''(''z'') is not a negative real multiple of ''f''(''z'') for any ''z'' = ''C''(''x''), thus 0 does not lie on the line segment joining ''f''(''C''(''x'')) to ''g''(''C''(''x'')), and
i.e., the [[winding number]] of the closed curve <math>f\circ C</math> around the origin; similarly for ''g''. The hypothesis ensures that ''g''(''z'') is not a negative real multiple of ''f''(''z'') for any ''z'' = ''C''(''x''), thus 0 does not lie on the line segment joining ''f''(''C''(''x'')) to ''g''(''C''(''x'')), and
:<math>H_t(x)=(1-t)f(C(x))+tg(C(x))</math>
<math display="block">H_t(x) = (1-t)f(C(x)) + t g(C(x))</math>
is a [[homotopy]] between the curves <math>f\circ C</math> and <math>g\circ C</math> avoiding the origin. The winding number is homotopy-invariant: the function
is a [[homotopy]] between the curves <math>f\circ C</math> and <math>g\circ C</math> avoiding the origin. The winding number is homotopy-invariant: the function
:<math>I(t)=\mathrm{Ind}_{H_t}(0)=\frac1{2\pi i}\oint_{H_t}\frac{dz}z</math>
<math display="block">I(t)=\mathrm{Ind}_{H_t}(0)=\frac1{2\pi i}\oint_{H_t}\frac{dz}z</math>
is continuous and integer-valued, hence constant. This shows
is continuous and integer-valued, hence constant. This shows
:<math>N_f(K)=\mathrm{Ind}_{f\circ C}(0)=\mathrm{Ind}_{g\circ C}(0)=N_g(K).</math>
<math display="block">N_f(K)=\mathrm{Ind}_{f\circ C}(0)=\mathrm{Ind}_{g\circ C}(0)=N_g(K).</math>


==See also==
== See also ==
* [[Fundamental theorem of algebra]], for its shortest demonstration yet, while using Rouché's theorem
* {{annotated link|Fundamental theorem of algebra}}
* [[Hurwitz's theorem (complex analysis)]]
* {{annotated link|Hurwitz's theorem (complex analysis)}}
* [[Rational root theorem]]
* {{annotated link|Rational root theorem}}
* [[Properties of polynomial roots]]
* {{annotated link|Properties of polynomial roots}}
* [[Riemann mapping theorem]]
* {{annotated link|Riemann mapping theorem}}
* [[Sturm's theorem]]
* {{annotated link|Sturm's theorem}}


==Notes==
== References ==
{{no footnotes|date=May 2015 }}
{{no footnotes|date=May 2015 }}
{{reflist}}
{{reflist}}
{{refbegin}}

*{{cite book|last=Beardon|first=Alan|title=Complex Analysis: The Argument Principle in Analysis and Topology|publisher=John Wiley and Sons|page=131|year=1979|isbn=0-471-99672-6}}
==References==
* {{cite book | first=Alan | last=Beardon | title=Complex Analysis: the Winding Number principle in analysis and topology | publisher=John Wiley and Sons | page=131 | year=1979 | isbn=0-471-99672-6 }}
* {{cite book | last=Conway | first=John B. | title=Functions of One Complex Variable I | publisher=Springer-Verlag New York | year=1978 | isbn=978-0-387-90328-6 }}
* {{cite book | first=E. C. | last=Titchmarsh | title=The Theory of Functions | edition=2nd | publisher=Oxford University Press | year=1939 | isbn=0-19-853349-7 | pages=117–119, 198–203 | authorlink=Edward Charles Titchmarsh }}
* {{cite book | last=Titchmarsh | first=E. C. | title=The Theory of Functions | url=https://archive.org/details/in.ernet.dli.2015.2588 | edition=2nd | publisher=Oxford University Press | year=1939 | isbn=0-19-853349-7 | pages=[https://archive.org/details/in.ernet.dli.2015.2588/page/n129 117]–119, 198–203 | authorlink=Edward Charles Titchmarsh }}
* Rouché É., ''Mémoire sur la série de Lagrange'', Journal de l'École Polytechnique, tome 22, 1862, p. 193-224. Theorem appears at p. 217. See [https://gallica.bnf.fr/ark:/12148/bpt6k433694t.r=%22Eugene%20Rouch%C3%A9%22?rk=21459;2 Gallica archives].

{{refend}}
==External links==
* [http://mathfaculty.fullerton.edu/mathews/c2003/RoucheTheoremMod.html Module for Rouche’s Theorem by John H. Mathews]


{{DEFAULTSORT:Rouche's theorem}}
{{DEFAULTSORT:Rouche's theorem}}
[[Category:Articles containing proofs]]
[[Category:Articles containing proofs]]

Latest revision as of 19:32, 4 December 2024

Rouché's theorem, named after Eugène Rouché, states that for any two complex-valued functions f and g holomorphic inside some region with closed contour , if |g(z)| < |f(z)| on , then f and f + g have the same number of zeros inside , where each zero is counted as many times as its multiplicity. This theorem assumes that the contour is simple, that is, without self-intersections. Rouché's theorem is an easy consequence of a stronger symmetric Rouché's theorem described below.

Usage

[edit]

The theorem is usually used to simplify the problem of locating zeros, as follows. Given an analytic function, we write it as the sum of two parts, one of which is simpler and grows faster than (thus dominates) the other part. We can then locate the zeros by looking at only the dominating part. For example, the polynomial has exactly 5 zeros in the disk since for every , and , the dominating part, has five zeros in the disk.

Geometric explanation

[edit]
As z travels along a closed curve C (not shown in the picture), f(z) and h(z) will trace out closed curves in the complex plane (shown in blue and red). So long as the curves never veer too far apart from each other (we require that f(z) remains closer to h(z) than the origin at all times), then the curves will wind around the origin the same number of times. Then, by the argument principle, f(z) and h(z) have the same number of zeros inside C (not shown).

It is possible to provide an informal explanation of Rouché's theorem.

Let C be a closed, simple curve (i.e., not self-intersecting). Let h(z) = f(z) + g(z). If f and g are both holomorphic on the interior of C, then h must also be holomorphic on the interior of C. Then, with the conditions imposed above, the Rouche's theorem in its original (and not symmetric) form says that

If |f(z)| > |h(z) − f(z)|, for every z in C, then f and h have the same number of zeros in the interior of C.

Notice that the condition |f(z)| > |h(z) − f(z)| means that for any z, the distance from f(z) to the origin is larger than the length of h(z) − f(z), which in the following picture means that for each point on the blue curve, the segment joining it to the origin is larger than the green segment associated with it. Informally we can say that the blue curve f(z) is always closer to the red curve h(z) than it is to the origin.

The previous paragraph shows that h(z) must wind around the origin exactly as many times as f(z). The index of both curves around zero is therefore the same, so by the argument principle, f(z) and h(z) must have the same number of zeros inside C.

One popular, informal way to summarize this argument is as follows: If a person were to walk a dog on a leash around and around a tree, such that the distance between the person and the tree is always greater than the length of the leash, then the person and the dog go around the tree the same number of times.

Applications

[edit]

Bounding roots

[edit]

Consider the polynomial with . By the quadratic formula it has two zeros at . Rouché's theorem can be used to obtain some hint about their positions. Since

Rouché's theorem says that the polynomial has exactly one zero inside the disk . Since is clearly outside the disk, we conclude that the zero is .

In general, a polynomial . If for some , then by Rouche's theorem, the polynomial has exactly roots inside .

This sort of argument can be useful in locating residues when one applies Cauchy's residue theorem.

Fundamental theorem of algebra

[edit]

Rouché's theorem can also be used to give a short proof of the fundamental theorem of algebra. Let and choose so large that: Since has zeros inside the disk (because ), it follows from Rouché's theorem that also has the same number of zeros inside the disk.

One advantage of this proof over the others is that it shows not only that a polynomial must have a zero but the number of its zeros is equal to its degree (counting, as usual, multiplicity).

Another use of Rouché's theorem is to prove the open mapping theorem for analytic functions. We refer to the article for the proof.

Symmetric version

[edit]

A stronger version of Rouché's theorem was published by Theodor Estermann in 1962.[1] It states: let be a bounded region with continuous boundary . Two holomorphic functions have the same number of roots (counting multiplicity) in , if the strict inequality holds on the boundary

The original version of Rouché's theorem then follows from this symmetric version applied to the functions together with the trivial inequality (in fact this inequality is strict since for some would imply ).

The statement can be understood intuitively as follows. By considering in place of , the condition can be rewritten as for . Since always holds by the triangle inequality, this is equivalent to saying that on , which in turn means that for the functions and are non-vanishing and .

Intuitively, if the values of and never pass through the origin and never point in the same direction as circles along , then and must wind around the origin the same number of times.

Proof of the symmetric form of Rouché's theorem

[edit]

Let be a simple closed curve whose image is the boundary . The hypothesis implies that f has no roots on , hence by the argument principle, the number Nf(K) of zeros of f in K is i.e., the winding number of the closed curve around the origin; similarly for g. The hypothesis ensures that g(z) is not a negative real multiple of f(z) for any z = C(x), thus 0 does not lie on the line segment joining f(C(x)) to g(C(x)), and is a homotopy between the curves and avoiding the origin. The winding number is homotopy-invariant: the function is continuous and integer-valued, hence constant. This shows

See also

[edit]

References

[edit]
  1. ^ Estermann, T. (1962). Complex Numbers and Functions. Athlone Press, Univ. of London. p. 156.