词条 | N-sphere |
释义 |
In mathematics, the {{math|n}}-sphere is the generalization of the ordinary sphere to spaces of arbitrary dimension. It is an {{math|n}}-dimensional manifold that can be embedded in Euclidean {{math|(n + 1)}}-space. The 0-sphere is a pair of points, the 1-sphere is a circle, and the 2-sphere is an ordinary sphere. Generally, when embedded in an {{math|(n + 1)}}-dimensional Euclidean space, an {{math|n}}-sphere is the surface or boundary of an {{math|(n + 1)}}-dimensional ball. That is, for any natural number {{math|n}}, an {{math|n}}-sphere of radius {{math|r}} may be defined in terms of an embedding in {{math|(n + 1)}}-dimensional Euclidean space as the set of points that are at distance {{math|r}} from a central point, where the radius {{math|r}} may be any positive real number. Thus, the {{math|n}}-sphere would be defined by: In particular:
An n-sphere embedded in an {{math|(n + 1)}}-dimensional Euclidean space is called a hypersphere. The {{math|n}}-sphere of unit radius is called the unit {{math|n}}-sphere, denoted {{math|Sn}}, often referred to as the {{math|n}}-sphere. When embedded as described, an {{math|n}}-sphere is the surface or boundary of an {{math|(n + 1)}}-dimensional ball. For {{math|n ≥ 2}}, the {{math|n}}-spheres are the simply connected {{math|n}}-dimensional manifolds of constant, positive curvature. The {{math|n}}-spheres admit several other topological descriptions: for example, they can be constructed by gluing two {{math|n}}-dimensional Euclidean spaces together, by identifying the boundary of an {{math|n}}-cube with a point, or (inductively) by forming the suspension of an {{math|(n − 1)}}-sphere. DescriptionFor any natural number {{math|n}}, an {{math|n}}-sphere of radius {{math|r}} is defined as the set of points in {{math|(n + 1)}}-dimensional Euclidean space that are at distance {{math|r}} from some fixed point {{math|c}}, where {{math|r}} may be any positive real number and where {{math|c}} may be any point in {{math|(n + 1)}}-dimensional space. In particular:
Euclidean coordinates in {{math|(n + 1)}}-spaceThe set of points in {{math|(n + 1)}}-space, {{math|(x1, x2, ..., xn+1)}}, that define an {{math|n}}-sphere, {{math|Sn}}, is represented by the equation: where {{math|c}}={{math|(c1, c2, ..., cn+1)}} is a center point, and {{math|r}} is the radius. The above {{math|n}}-sphere exists in {{math|(n + 1)}}-dimensional Euclidean space and is an example of an {{math|n}}-manifold. The volume form {{math|ω}} of an {{math|n}}-sphere of radius {{math|r}} is given by where {{math|∗}} is the Hodge star operator; see {{harvtxt|Flanders|1989|loc=§6.1}} for a discussion and proof of this formula in the case {{math|r {{=}} 1}}. As a result, === {{math|n}}-ball === {{main|Ball (mathematics)}}The space enclosed by an {{math|n}}-sphere is called an {{math|(n + 1)}}-ball. An {{math|(n + 1)}}-ball is closed if it includes the {{math|n}}-sphere, and it is open if it does not include the {{math|n}}-sphere. Specifically:
Topological descriptionTopologically, an {{math|n}}-sphere can be constructed as a one-point compactification of {{math|n}}-dimensional Euclidean space. Briefly, the {{math|n}}-sphere can be described as {{math|Sn {{=}} Rn ∪ {∞}}}, which is {{math|n}}-dimensional Euclidean space plus a single point representing infinity in all directions. In particular, if a single point is removed from an {{math|n}}-sphere, it becomes homeomorphic to {{math|Rn}}. This forms the basis for stereographic projection.[1] Volume and surface area{{see also|Volume of an n-ball}}{{math|Vn(R)}} and {{math|Sn(R)}} are the {{math|n}}-dimensional volume of the {{math|n}}-ball and the surface area of the {{math|n}}-sphere embedded in dimension {{math|n + 1}}, respectively, of radius {{math|R}}. {{hypersphere_volume_and_surface_area_graphs.svg}}The constants {{math|Vn}} and {{math|Sn}} (for {{math|R {{=}} 1}}, the unit ball and sphere) are related by the recurrences: The surfaces and volumes can also be given in closed form: where {{math|Γ}} is the gamma function. Derivations of these equations are given in this section. In general, the volume of the {{math|n}}-ball in {{math|n}}-dimensional Euclidean space, and the surface area of the {{math|n}}-sphere in {{math|(n + 1)}}-dimensional Euclidean space, of radius {{math|R}}, are proportional to the {{math|n}}th power of the radius, {{math|R}} (with different constants of proportionality that vary with {{math|n}}). We write {{math|Vn(R) {{=}} VnRn}} for the volume of the {{math|n}}-ball and {{math|Sn(R) {{=}} SnRn}} for the surface area of the {{math|n}}-sphere, both of radius {{math|R}}, where {{math|Vn {{=}} Vn(1)}} and {{math|Sn {{=}} Sn(1)}} are the values for the unit-radius case. Given the radius {{math|R}}, the enclosed volume and the surface area of the {{math|n}}-sphere reach a maximum and then decrease towards zero as the dimension {{math|n}} increases. In particular, the volume {{math|Vn(R)}} enclosed by the {{math|(n – 1)}}-sphere of constant radius {{math|R}} embedded in {{math|n}} dimensions reaches a maximum at the dimension {{math|n}} satisfying {{math|Vn−1 < VnR}} and {{math|Vn ≥ Vn+1R}} where {{math|Vn}} is given in the sidebar to the right; if the last (weak) inequality holds with equality, then the same maximum also occurs at {{math|n + 1}}. Specifically, for unit radius the largest enclosed volume is that enclosed by a 4-dimensional sphere bounding a 5-dimensional ball. Similarly, the surface area {{math|Sn(R)}} of the {{mvar|n}}-sphere of constant radius {{mvar|R}} embedded in {{mvar|n + 1}} dimensions reaches a maximum for dimension {{math|n}} that satisfies {{math|Sn−1 < SnR}} and {{math|Sn ≥ Sn+1R}} where {{math|Sn}} is given in the sidebar to the right; if the last (weak) inequality holds with equality, then the same maximum also occurs at {{math|n + 1}}.[2] Specifically, for unit radius the largest surface area occurs for the 6-dimensional sphere bounding a 7-dimensional ball. ExamplesThe 0-ball consists of a single point. The 0-dimensional Hausdorff measure is the number of points in a set. So, The unit 1-ball is the interval {{math|[−1,1]}} of length 2. So, The 0-sphere consists of its two end-points, {{math|{−1,1}}}. So, The unit 1-sphere is the unit circle in the Euclidean plane, and this has circumference (1-dimensional measure) The region enclosed by the unit 1-sphere is the 2-ball, or unit disc, and this has area (2-dimensional measure) Analogously, in 3-dimensional Euclidean space, the surface area (2-dimensional measure) of the unit 2-sphere is given by and the volume enclosed is the volume (3-dimensional measure) of the unit 3-ball, given by RecurrencesThe surface area, or properly the {{math|n}}-dimensional volume, of the {{math|n}}-sphere at the boundary of the {{math|(n + 1)}}-ball of radius {{math|R}} is related to the volume of the ball by the differential equation or, equivalently, representing the unit {{math|n}}-ball as a union of concentric {{math|(n − 1)}}-sphere shells, So, We can also represent the unit {{math|(n + 2)}}-sphere as a union of tori, each the product of a circle (1-sphere) with an {{math|n}}-sphere. Let {{math|r {{=}} cos θ}} and {{math|r2 + R2 {{=}} 1}}, so that {{math|R {{=}} sin θ}} and {{math|dR {{=}} cos θ dθ}}. Then, Since {{math|S1 {{=}} 2π V0}}, the equation holds for all {{math|n}}. This completes our derivation of the recurrences: Closed formsCombining the recurrences, we see that So it is simple to show by induction on k that, where {{math|!!}} denotes the double factorial, defined for odd integers {{math|2k + 1}} by {{math|1=(2k + 1)!! = 1 × 3 × 5 ... (2k − 1) × (2k + 1)}}. In general, the volume, in {{math|n}}-dimensional Euclidean space, of the unit {{math|n}}-ball, is given by where {{math|Γ}} is the gamma function, which satisfies {{math|Γ({{sfrac|1|2}}) {{=}} {{sqrt|π}}}}, {{math|Γ(1) {{=}} 1}}, and {{math|Γ(x + 1) {{=}} xΓ(x)}}. By multiplying {{math|Vn}} by {{math|Rn}}, differentiating with respect to {{math|R}}, and then setting {{math|R {{=}} 1}}, we get the closed form Other relationsThe recurrences can be combined to give a "reverse-direction" recurrence relation for surface area, as depicted in the diagram: Index-shifting {{math|n}} to {{math|n − 2}} then yields the recurrence relations: where {{math|S0 {{=}} 2}}, {{math|V1 {{=}} 2}}, {{math|S1 {{=}} 2π}} and {{math|V2 {{=}} π}}. The recurrence relation for {{math|Vn}} can also be proved via integration with 2-dimensional polar coordinates: ==Spherical coordinates== We may define a coordinate system in an {{math|n}}-dimensional Euclidean space which is analogous to the spherical coordinate system defined for 3-dimensional Euclidean space, in which the coordinates consist of a radial coordinate {{math|r}}, and {{math|n − 1}} angular coordinates {{math|φ1, φ2, ... φn−1}}, where the angles {{math|φ1, φ2, ... φn−2}} range over {{math|[0,π]}} radians (or over {{math|[0,180]}} degrees) and {{math|φn−1}} ranges over {{math|[0,2π)}} radians (or over {{math|[0,360)}} degrees). If {{math|xi}} are the Cartesian coordinates, then we may compute {{math|x1, ... xn}} from {{math|r, φ1, ... φn−1}} with: [3] Except in the special cases described below, the inverse transformation is unique: where if {{math|xk ≠ 0}} for some {{math|k}} but all of {{math|xk+1, ... xn}} are zero then {{math|φk {{=}} 0}} when {{math|xk > 0}}, and {{math|φk {{=}} π}} (180 degrees) when {{math|xk < 0}}. There are some special cases where the inverse transform is not unique; {{math|φk}} for any {{math|k}} will be ambiguous whenever all of {{math|xk, xk+1, ... xn}} are zero; in this case {{math|φk}} may be chosen to be zero. Spherical volume elementExpressing the angular measures in radians, the volume element in {{math|n}}-dimensional Euclidean space will be found from the Jacobian of the transformation: and the above equation for the volume of the {{math|n}}-ball can be recovered by integrating: The volume element of the {{math|(n − 1)}}-sphere, which generalizes the area element of the 2-sphere, is given by The natural choice of an orthogonal basis over the angular coordinates is a product of ultraspherical polynomials, for {{math|j {{=}} 1, 2,... n − 2}}, and the {{math|eisφj}} for the angle {{math|j {{=}} n − 1}} in concordance with the spherical harmonics. Stereographic projection{{main|Stereographic projection}}Just as a two-dimensional sphere embedded in three dimensions can be mapped onto a two-dimensional plane by a stereographic projection, an {{math|n}}-sphere can be mapped onto an {{math|n}}-dimensional hyperplane by the {{math|n}}-dimensional version of the stereographic projection. For example, the point {{math|[x,y,z]}} on a two-dimensional sphere of radius 1 maps to the point {{math|[{{sfrac|x|1 − z}},{{sfrac|y|1 − z}}]}} on the {{math|xy}}-plane. In other words, Likewise, the stereographic projection of an {{math|n}}-sphere {{math|Sn−1}} of radius 1 will map to the {{math|(n − 1)}}-dimensional hyperplane {{math|Rn−1}} perpendicular to the {{math|xn}}-axis as Generating random pointsUniformly at random on the {{math|(n − 1)}}-sphereTo generate uniformly distributed random points on the unit {{math|(n − 1)}}-sphere (that is, the surface of the unit {{math|n}}-ball), {{harvtxt|Marsaglia|1972}} gives the following algorithm. Generate an {{math|n}}-dimensional vector of normal deviates (it suffices to use {{math|N(0, 1)}}, although in fact the choice of the variance is arbitrary), {{math|x {{=}} (x1, x2,... xn)}}. Now calculate the "radius" of this point: The vector {{math|{{sfrac|1|r}}x}} is uniformly distributed over the surface of the unit {{math|n}}-ball. An alternative given by Marsaglia is to uniformly randomly select a point {{math|x {{=}} (x1, x2,... xn)}} in the unit {{math|n}}-cube by sampling each {{math|xi}} independently from the uniform distribution over {{math|(–1,1)}}, computing {{math|r}} as above, and rejecting the point and resampling if {{math|r ≥ 1}} (i.e., if the point is not in the {{math|n}}-ball), and when a point in the ball is obtained scaling it up to the spherical surface by the factor {{math|{{sfrac|1|r}}}}; then again {{math|{{sfrac|1|r}}x}} is uniformly distributed over the surface of the unit {{math|n}}-ball. Uniformly at random within the {{math|n}}-ballWith a point selected uniformly at random from the surface of the unit {{math|(n - 1)}}-sphere (e.g., by using Marsaglia's algorithm), one needs only a radius to obtain a point uniformly at random from within the unit {{math|n}}-ball. If {{math|u}} is a number generated uniformly at random from the interval {{math|[0, 1]}} and {{math|x}} is a point selected uniformly at random from the unit {{math|(n - 1)}}-sphere, then {{math|u{{frac|1|n}}x}} is uniformly distributed within the unit {{math|n}}-ball. Alternatively, points may be sampled uniformly from within the unit {{math|n}}-ball by a reduction from the unit {{math|(n + 1)}}-sphere. In particular, if {{math|(x1,x2,...,xn+2)}} is a point selected uniformly from the unit {{math|(n + 1)}}-sphere, then {{math|(x1,x2,...,xn)}} is uniformly distributed within the unit {{math|n}}-ball (i.e., by simply discarding two coordinates).[4] Note that if {{math|n}} is sufficiently large, most of the volume of the {{math|n}}-ball will be contained in the region very close to its surface, so a point selected from that volume will also probably be close to the surface. This is one of the phenomena leading to the so-called curse of dimensionality that arises in some numerical and other applications. Specific spheres
.
See also{{div col|colwidth=15em}}
Notes1. ^James W. Vick (1994). Homology theory, p. 60. Springer 2. ^{{cite journal|last1=Loskot|first1=Pavel|title=On Monotonicity of the Hypersphere Volume and Area|journal=Journal of Geometry|date=November 2007|volume=87|issue=1–2|pages=96–98|doi=10.1007/s00022-007-1891-1}} 3. ^{{cite journal |last1=Blumenson |first1=L. E. |title=A Derivation of n-Dimensional Spherical Coordinates |journal=The American Mathematical Monthly |date=1960 |volume=67 |issue=1 |pages=63–66 |jstor=2308932 |doi=10.2307/2308932 }} 4. ^{{cite report|first1=Aaron R. | last1=Voelker | first2=Jan | last2=Gosmann | first3=Terrence C. | last3=Stewart | title=Efficiently sampling vectors and coordinates from the n-sphere and n-ball | year=2017 | publisher=Centre for Theoretical Neuroscience | url=http://compneuro.uwaterloo.ca/publications/voelker2017.html | doi=10.13140/RG.2.2.15829.01767/1}} References
}}
|year=1999 | journal = Phys. Rev. A | doi=10.1103/PhysRevA.59.1135 |volume=59 | number=2|pages=1135–1146}} External links
2 : Multi-dimensional geometry|Spheres |
随便看 |
|
开放百科全书收录14589846条英语、德语、日语等多语种百科知识,基本涵盖了大多数领域的百科知识,是一部内容自由、开放的电子版国际百科全书。