Movatterモバイル変換


[0]ホーム

URL:


Jump to content
WikipediaThe Free Encyclopedia
Search

Quaternionic analysis

From Wikipedia, the free encyclopedia
Function theory with quaternion variable

Inmathematics,quaternionic analysis is the study offunctions withquaternions as thedomain and/or range. Such functions can be calledfunctions of a quaternion variable just asfunctions of a real variable or acomplex variable are called.

As withcomplex andreal analysis, it is possible to study the concepts ofanalyticity,holomorphy,harmonicity andconformality in the context of quaternions. Unlike thecomplex numbers and like thereals, the four notions do not coincide.

Properties

[edit]

Theprojections of a quaternion onto its scalar part or onto its vector part, as well as the modulus andversor functions, are examples that are basic to understanding quaternion structure.

An important example of a function of a quaternion variable is

f1(q)=uqu1{\displaystyle f_{1}(q)=uqu^{-1}}

whichrotates the vector part ofq by twice the angle represented by the versoru.

The quaternionmultiplicative inversef2(q)=q1{\displaystyle f_{2}(q)=q^{-1}} is another fundamental function, but as with other number systems,f2(0){\displaystyle f_{2}(0)} and related problems are generally excluded due to the nature ofdividing by zero.

Affine transformations of quaternions have the form

f3(q)=aq+b,a,b,qH.{\displaystyle f_{3}(q)=aq+b,\quad a,b,q\in \mathbb {H} .}

Linear fractional transformations of quaternions can be represented by elements of thematrix ringM2(H){\displaystyle M_{2}(\mathbb {H} )} operating on theprojective line overH{\displaystyle \mathbb {H} }. For instance, the mappingsquqv,{\displaystyle q\mapsto uqv,} whereu{\displaystyle u} andv{\displaystyle v} are fixedversors serve to produce themotions of elliptic space.

Quaternion variable theory differs in some respects from complex variable theory. For example: Thecomplex conjugate mapping of the complex plane is a central tool but requires the introduction of a non-arithmetic,non-analytic operation. Indeed, conjugation changes theorientation of plane figures, something that arithmetic functions do not change.

In contrast to thecomplex conjugate, the quaternion conjugation can be expressed arithmetically, asf4(q)=12(q+iqi+jqj+kqk){\displaystyle f_{4}(q)=-{\tfrac {1}{2}}(q+iqi+jqj+kqk)}

This equation can be proven, starting with thebasis {1, i, j, k}:

f4(1)=12(1111)=1,f4(i)=12(ii+i+i)=i,f4(j)=j,f4(k)=k{\displaystyle f_{4}(1)=-{\tfrac {1}{2}}(1-1-1-1)=1,\quad f_{4}(i)=-{\tfrac {1}{2}}(i-i+i+i)=-i,\quad f_{4}(j)=-j,\quad f_{4}(k)=-k}.

Consequently, sincef4{\displaystyle f_{4}} islinear,

f4(q)=f4(w+xi+yj+zk)=wf4(1)+xf4(i)+yf4(j)+zf4(k)=wxiyjzk=q.{\displaystyle f_{4}(q)=f_{4}(w+xi+yj+zk)=wf_{4}(1)+xf_{4}(i)+yf_{4}(j)+zf_{4}(k)=w-xi-yj-zk=q^{*}.}

The success ofcomplex analysis in providing a rich family ofholomorphic functions for scientific work has engaged some workers in efforts to extend the planar theory, based on complex numbers, to a 4-space study with functions of a quaternion variable.[1] These efforts were summarized inDeavours (1973).[a]

ThoughH{\displaystyle \mathbb {H} }appears as a union of complex planes, the following proposition shows that extending complex functions requires special care:

Letf5(z)=u(x,y)+iv(x,y){\displaystyle f_{5}(z)=u(x,y)+iv(x,y)} be a function of a complex variable,z=x+iy{\displaystyle z=x+iy}. Suppose also thatu{\displaystyle u} is aneven function ofy{\displaystyle y} and thatv{\displaystyle v} is anodd function ofy{\displaystyle y}. Thenf5(q)=u(x,y)+rv(x,y){\displaystyle f_{5}(q)=u(x,y)+rv(x,y)} is an extension off5{\displaystyle f_{5}} to a quaternion variableq=x+yr{\displaystyle q=x+yr} wherer2=1{\displaystyle r^{2}=-1} andrH{\displaystyle r\in \mathbb {H} }.Then, letr{\displaystyle r^{*}} represent the conjugate ofr{\displaystyle r}, so thatq=xyr{\displaystyle q=x-yr^{*}}. The extension toH{\displaystyle \mathbb {H} } will be complete when it is shown thatf5(q)=f5(xyr){\displaystyle f_{5}(q)=f_{5}(x-yr^{*})}. Indeed, by hypothesis

u(x,y)=u(x,y),v(x,y)=v(x,y){\displaystyle u(x,y)=u(x,-y),\quad v(x,y)=-v(x,-y)\quad } one obtains
f5(xyr)=u(x,y)+rv(x,y)=u(x,y)+rv(x,y)=f5(q).{\displaystyle f_{5}(x-yr^{*})=u(x,-y)+r^{*}v(x,-y)=u(x,y)+rv(x,y)=f_{5}(q).}

Homographies

[edit]

In the following, colons and square brackets are used to denotehomogeneous vectors.

Therotation about axisr is a classical application of quaternions tospace mapping.[2]In terms of ahomography, the rotation is expressed

[q:1](u00u)=[qu:u][u1qu:1],{\displaystyle [q:1]{\begin{pmatrix}u&0\\0&u\end{pmatrix}}=[qu:u]\thicksim [u^{-1}qu:1],}

whereu=exp(θr)=cosθ+rsinθ{\displaystyle u=\exp(\theta r)=\cos \theta +r\sin \theta } is aversor. Ifp * = −p, then the translationqq+p{\displaystyle q\mapsto q+p} is expressed by

[q:1](10p1)=[q+p:1].{\displaystyle [q:1]{\begin{pmatrix}1&0\\p&1\end{pmatrix}}=[q+p:1].}

Rotation and translationxr along the axis of rotation is given by

[q:1](u0uxru)=[qu+uxr:u][u1qu+xr:1].{\displaystyle [q:1]{\begin{pmatrix}u&0\\uxr&u\end{pmatrix}}=[qu+uxr:u]\thicksim [u^{-1}qu+xr:1].}

Such a mapping is called ascrew displacement. In classicalkinematics,Chasles' theorem states that any rigid body motion can be displayed as a screw displacement. Just as the representation of aEuclidean plane isometry as a rotation is a matter of complex number arithmetic, so Chasles' theorem, and thescrew axis required, is a matter of quaternion arithmetic with homographies: Lets be a right versor, or square root of minus one, perpendicular tor, witht =rs.

Consider the axis passing throughs and parallel tor. Rotation about it is expressed[3] by the homography composition

(10s1)(u00u)(10s1)=(u0zu),{\displaystyle {\begin{pmatrix}1&0\\-s&1\end{pmatrix}}{\begin{pmatrix}u&0\\0&u\end{pmatrix}}{\begin{pmatrix}1&0\\s&1\end{pmatrix}}={\begin{pmatrix}u&0\\z&u\end{pmatrix}},}

wherez=ussu=sinθ(rssr)=2tsinθ.{\displaystyle z=us-su=\sin \theta (rs-sr)=2t\sin \theta .}

Now in the (s,t)-plane the parameter θ traces out a circleu1z=u1(2tsinθ)=2sinθ(tcosθssinθ){\displaystyle u^{-1}z=u^{-1}(2t\sin \theta )=2\sin \theta (t\cos \theta -s\sin \theta )} in the half-plane{wt+xs:x>0}.{\displaystyle \lbrace wt+xs:x>0\rbrace .}

Anyp in this half-plane lies on a ray from the origin through the circle{u1z:0<θ<π}{\displaystyle \lbrace u^{-1}z:0<\theta <\pi \rbrace } and can be writtenp=au1z,  a>0.{\displaystyle p=au^{-1}z,\ \ a>0.}

Thenup =az, with(u0azu){\displaystyle {\begin{pmatrix}u&0\\az&u\end{pmatrix}}} as the homography expressingconjugation of a rotation by a translation p.

The derivative for quaternions

[edit]

Since the time of Hamilton, it has been realized that requiring the independence of thederivative from the path that a differential follows toward zero is too restrictive: it excludes even f(q)=q2 {\displaystyle \ f(q)=q^{2}\ } from differentiation. Therefore, a direction-dependent derivative is necessary for functions of a quaternion variable.[4][5]Considering the increment ofpolynomial function of quaternionic argument shows that the increment is a linear map of increment of the argument.[dubiousdiscuss] From this, a definition can be made:

A continuous function f:HH {\displaystyle \ f:\mathbb {H} \rightarrow \mathbb {H} \ }is calleddifferentiable on the set UH ,{\displaystyle \ U\subset \mathbb {H} \ ,} if at every point xU ,{\displaystyle \ x\in U\ ,} an increment of the function f {\displaystyle \ f\ } corresponding to a quaternion increment h {\displaystyle \ h\ } of its argument, can be represented as

f(x+h)f(x)=df(x)dxh+o(h){\displaystyle f(x+h)-f(x)={\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\circ h+o(h)}

where

df(x)dx:HH{\displaystyle {\frac {\operatorname {d} f(x)}{\operatorname {d} x}}:\mathbb {H} \rightarrow \mathbb {H} }

islinear map of quaternion algebra H ,{\displaystyle \ \mathbb {H} \ ,} and o:HH {\displaystyle \ o:\mathbb {H} \rightarrow \mathbb {H} \ }represents some continuous map such that

lima0 | o(a) | | a |=0 ,{\displaystyle \lim _{a\rightarrow 0}{\frac {\ \left|\ o(a)\ \right|\ }{\left|\ a\ \right|}}=0\ ,}

and the notation h {\displaystyle \ \circ h\ } denotes ...[further explanation needed]

The linear mapdf(x)dx{\displaystyle {\frac {\operatorname {d} f(x)}{\operatorname {d} x}}}is called the derivative of the map f .{\displaystyle \ f~.}

On the quaternions, the derivative may be expressed as

df(x)dx=sds0f(x)dxds1f(x)dx{\displaystyle {\frac {\operatorname {d} f(x)}{\operatorname {d} x}}=\sum _{s}{\frac {\operatorname {d} _{s0}f(x)}{\operatorname {d} x}}\otimes {\frac {\operatorname {d} _{s1}f(x)}{\operatorname {d} x}}}

Therefore, the differential of the map f {\displaystyle \ f\ } may be expressed as follows, with brackets on either side.

df(x)dxdx=(sds0f(x)dxds1f(x)dx)dx=sds0f(x)dx(dx)ds1f(x)dx{\displaystyle {\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\circ \operatorname {d} x=\left(\sum _{s}{\frac {\operatorname {d} _{s0}f(x)}{\operatorname {d} x}}\otimes {\frac {\operatorname {d} _{s1}f(x)}{\operatorname {d} x}}\right)\circ \operatorname {d} x=\sum _{s}{\frac {\operatorname {d} _{s0}f(x)}{\operatorname {d} x}}\left(\operatorname {d} x\right){\frac {\operatorname {d} _{s1}f(x)}{\operatorname {d} x}}}

The number of terms in the sum will depend on the function f .{\displaystyle \ f~.} The expressions  dspdf(x)dx   for   p=0,1  {\displaystyle ~~{\frac {\operatorname {d} _{sp}\operatorname {d} f(x)}{\operatorname {d} x}}~~{\mathsf {\ for\ }}~~p=0,1~~} are calledcomponents of derivative.

The derivative of a quaternionic function is defined by the expression

df(x)dxh=limt0(  f(x+t h)f(x) t ){\displaystyle {\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\circ h=\lim _{t\to 0}\left(\ {\frac {\ f(x+t\ h)-f(x)\ }{t}}\ \right)}

where the variable t {\displaystyle \ t\ } is a real scalar.

The following equations then hold:

d(f(x)+g(x))dx=df(x)dx+dg(x)dx{\displaystyle {\frac {\operatorname {d} \left(f(x)+g(x)\right)}{\operatorname {d} x}}={\frac {\operatorname {d} f(x)}{\operatorname {d} x}}+{\frac {\operatorname {d} g(x)}{\operatorname {d} x}}}
d(f(x) g(x))dx=df(x)dx g(x)+f(x) dg(x)dx{\displaystyle {\frac {\operatorname {d} \left(f(x)\ g(x)\right)}{\operatorname {d} x}}={\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\ g(x)+f(x)\ {\frac {\operatorname {d} g(x)}{\operatorname {d} x}}}
d(f(x) g(x))dxh=(df(x)dxh) g(x)+f(x)(dg(x)dxh){\displaystyle {\frac {\operatorname {d} \left(f(x)\ g(x)\right)}{\operatorname {d} x}}\circ h=\left({\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\circ h\right)\ g(x)+f(x)\left({\frac {\operatorname {d} g(x)}{\operatorname {d} x}}\circ h\right)}
d(a f(x) b)dx=a df(x)dx b{\displaystyle {\frac {\operatorname {d} \left(a\ f(x)\ b\right)}{\operatorname {d} x}}=a\ {\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\ b}
d(a f(x) b)dxh=a(df(x)dxh)b{\displaystyle {\frac {\operatorname {d} \left(a\ f(x)\ b\right)}{\operatorname {d} x}}\circ h=a\left({\frac {\operatorname {d} f(x)}{\operatorname {d} x}}\circ h\right)b}

For the function f(x)=a x b ,{\displaystyle \ f(x)=a\ x\ b\ ,} where a {\displaystyle \ a\ } and b {\displaystyle \ b\ } are constant quaternions, the derivative is

d(a x b)dx=ab{\displaystyle {\frac {\operatorname {d} \left(a\ x\ b\right)}{\operatorname {d} x}}=a\otimes b}dy=d(a x b)dxdx=a (dx) b{\displaystyle \operatorname {d} y={\frac {\operatorname {d} \left(a\ x\ b\right)}{\operatorname {d} x}}\circ \operatorname {d} x=a\ \left(\operatorname {d} x\right)\ b}

and so the components are:

d10(a x b)dx=a{\displaystyle {\frac {\operatorname {d} _{10}\left(a\ x\ b\right)}{\operatorname {d} x}}=a}d11(a x b)dx=b{\displaystyle {\frac {\operatorname {d} _{11}\left(a\ x\ b\right)}{\operatorname {d} x}}=b}

Similarly, for the function f(x)=x2 ,{\displaystyle \ f(x)=x^{2}\ ,} the derivative is

dx2dx=x1+1x{\displaystyle {\frac {\operatorname {d} x^{2}}{\operatorname {d} x}}=x\otimes 1+1\otimes x}dy=dx2dxdx=x dx+(dx) x{\displaystyle \operatorname {d} y={\frac {\operatorname {d} x^{2}}{\operatorname {d} x}}\circ \operatorname {d} x=x\ \operatorname {d} x+(\operatorname {d} x)\ x}

and the components are:

d10x2dx=x{\displaystyle {\frac {\operatorname {d} _{10}x^{2}}{\operatorname {d} x}}=x}d11x2dx=1{\displaystyle {\frac {\operatorname {d} _{11}x^{2}}{\operatorname {d} x}}=1}
d20x2dx=1{\displaystyle {\frac {\operatorname {d} _{20}x^{2}}{\operatorname {d} x}}=1}d21x2dx=x{\displaystyle {\frac {\operatorname {d} _{21}x^{2}}{\operatorname {d} x}}=x}

Finally, for the function f(x)=x1 ,{\displaystyle \ f(x)=x^{-1}\ ,} the derivative is

dx1dx=x1x1{\displaystyle {\frac {\operatorname {d} x^{-1}}{\operatorname {d} x}}=-x^{-1}\otimes x^{-1}}dy=dx1dxdx=x1(dx) x1{\displaystyle \operatorname {d} y={\frac {\operatorname {d} x^{-1}}{\operatorname {d} x}}\circ \operatorname {d} x=-x^{-1}(\operatorname {d} x)\ x^{-1}}

and the components are:

d10x1dx=x1{\displaystyle {\frac {\operatorname {d} _{10}x^{-1}}{\operatorname {d} x}}=-x^{-1}}d11x1dx=x1{\displaystyle {\frac {\operatorname {d} _{11}x^{-1}}{\operatorname {d} x}}=x^{-1}}

See also

[edit]

Notes

[edit]
  1. ^Deavours (1973) recalls a 1935 issue ofCommentarii Mathematici Helvetici where an alternative theory of "regular functions" was initiated byFueter (1936) through the idea ofMorera's theorem: quaternion functionF{\displaystyle F} is "left regular atq{\displaystyle q}" when the integral ofF{\displaystyle F} vanishes over any sufficiently smallhypersurface containingq{\displaystyle q}. Then the analogue ofLiouville's theorem holds: The only regular quaternion function with bounded norm inE4{\displaystyle \mathbb {E} ^{4}} is a constant. One approach to construct regular functions is to usepower series with real coefficients. Deavours also gives analogues for thePoisson integral, theCauchy integral formula, and the presentation ofMaxwell’s equations of electromagnetism with quaternion functions.

Citations

[edit]
  1. ^(Fueter 1936)
  2. ^(Cayley 1848, especially page 198)
  3. ^(Hamilton 1853, §287 pp. 273,4)
  4. ^Hamilton (1866), Chapter II, On differentials and developments of functions of quaternions, pp. 391–495
  5. ^Laisant (1881), Chapitre 5: Différentiation des Quaternions, pp. 104–117

References

[edit]
Basic concepts
Derivatives
Measurability
Integrals
Results
Related
Functional calculus
Applications
Retrieved from "https://en.wikipedia.org/w/index.php?title=Quaternionic_analysis&oldid=1324063939"
Categories:
Hidden categories:

[8]ページ先頭

©2009-2026 Movatter.jp