{
  "id": "parameterization",
  "title": "Parameterization",
  "lang": "en",
  "source_lang": "en",
  "summary": "Parameterization represents a geometric object, mathematical family, or model using parameters whose values determine its points or members.",
  "category": "mathematics",
  "subcategory": "geometry",
  "type": "concept",
  "aliases": [
    "Parametrization",
    "Parametric representation"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/parameterization/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/parameterization/",
      "json": "https://aiwiki.page/zh/parameterization.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Parameterization is the representation of a mathematical object or family of objects by quantities called parameters. In geometry, parameters specify points on a curve, surface, or higher-dimensional space; in statistics, they identify members of a family of probability distributions. A parameterization typically takes the form of a map from a parameter space to the objects being represented. Different parameterizations can describe the same object while differing in uniqueness, smoothness, computational convenience, and interpretation."
    },
    {
      "heading": "General formulation",
      "text": "A parameterization can be expressed as a function\n\n\\[F:\\Theta\\longrightarrow X,\\]\n\nwhere \\(\\Theta\\) is the parameter space and \\(X\\) is the space containing the objects of interest. A parameter value \\(\\theta\\in\\Theta\\) determines the object \\(F(\\theta)\\). The represented family is the image\n\n\\[F(\\Theta)=\\{F(\\theta):\\theta\\in\\Theta\\}.\\]\n\nThe parameter space may be an interval, a subset of Euclidean space, a discrete set, or a more general space. Parameters need not be real numbers, although finite lists of real-valued parameters are especially common.\n\nSeveral properties must be distinguished:\n\n- Coverage: the image includes every intended point or object.\n- Uniqueness: different parameter values represent different objects.\n- Regularity: the map has suitable continuity or differentiability properties.\n- Minimality: the representation avoids unnecessary parameters, relative to the structure being studied.\n\nA parameterization need not satisfy all these properties. For example, a periodic representation may deliberately assign many parameter values to the same point.\n\nThe distinction between a parameter and a variable depends on context. In a family \\(y=f(x;\\theta)\\), \\(x\\) usually varies within an individual function, while \\(\\theta\\) selects which function in the family is being considered. When studying the whole family, \\(\\theta\\) itself becomes a variable."
    },
    {
      "heading": "Curves and parametric equations",
      "text": "A curve in \\(\\mathbb R^n\\) is commonly represented by a map\n\n\\[\\mathbf r:I\\longrightarrow\\mathbb R^n, \\qquad t\\longmapsto\\mathbf r(t),\\]\n\nwhere \\(I\\) is an interval. In three dimensions, this becomes a system of parametric equations:\n\n\\[x=x(t),\\qquad y=y(t),\\qquad z=z(t).\\]\n\nThe parameter \\(t\\) may represent time, but it need not have a physical interpretation.\n\nFor example, the unit circle has the parameterization\n\n\\[\\mathbf r(t)=(\\cos t,\\sin t).\\]\n\nTaking \\(0\\leq t\\leq2\\pi\\) covers the circle but represents \\((1,0)\\) twice. Taking \\(0\\leq t<2\\pi\\) makes the representation one-to-one, but its inverse is not continuous at the point where the angular coordinate wraps around. Thus a global one-to-one representation need not be a global coordinate system compatible with the topology.\n\nParametric, explicit, and implicit descriptions\n\nThe same geometric set may admit several forms of description:\n\n- Explicit: \\(y=f(x)\\).\n- Implicit: \\(G(x,y)=0\\).\n- Parametric: \\(x=x(t)\\), \\(y=y(t)\\).\n\nFor the unit circle, the implicit equation is\n\n\\[x^2+y^2=1.\\]\n\nNo single expression \\(y=f(x)\\) describes the whole circle, whereas a parametric representation does.\n\nAn explicit graph is itself parameterized by\n\n\\[\\mathbf r(t)=(t,f(t)).\\]\n\nParameterization therefore includes graph representations but also handles vertical tangents, closed curves, and curves that cross themselves.\n\nRegularity and singularities\n\nFor a differentiable curve, the derivative \\(\\mathbf r'(t)\\) gives a tangent vector. A parameterization is regular at \\(t\\) when\n\n\\[\\mathbf r'(t)\\ne\\mathbf0.\\]\n\nFailure of regularity can arise from the representation rather than from the geometric set. For instance,\n\n\\[\\mathbf r(t)=(t^3,0)\\]\n\ntraces a straight line, yet its derivative vanishes at \\(t=0\\). A singularity of a parameterization must therefore be distinguished from an intrinsic singularity of the represented object."
    },
    {
      "heading": "Reparameterization and arc length",
      "text": "A reparameterization changes the parameters without changing the represented geometric image. If\n\n\\[\\mathbf r:I\\to\\mathbb R^n\\]\n\nand \\(\\phi:J\\to I\\) is a suitable bijection, then\n\n\\[\\widetilde{\\mathbf r}(u)=\\mathbf r(\\phi(u))\\]\n\nrepresents the same curve.\n\nFor regular differentiable reparameterizations, \\(\\phi\\) and its inverse are differentiable and \\(\\phi'\\) is nonzero. The chain rule gives\n\n\\[\\widetilde{\\mathbf r}'(u) =\\mathbf r'(\\phi(u))\\phi'(u).\\]\n\nAn increasing \\(\\phi\\) preserves orientation; a decreasing one reverses it. The speed changes by the factor \\(|\\phi'(u)|\\).\n\nFor a continuously differentiable regular curve, arc length from a reference parameter \\(t_0\\) is\n\n\\[s(t)=\\int_{t_0}^{t}\\|\\mathbf r'(\\tau)\\|\\,d\\tau.\\]\n\nWhere this relation can be inverted, writing the curve in terms of \\(s\\) produces a unit-speed parameterization:\n\n\\[\\left\\|\\frac{d\\mathbf r}{ds}\\right\\|=1.\\]\n\nThis separates the curve’s geometry from the rate at which the original parameter traverses it. For a physical trajectory, however, replacing time by arc length preserves the spatial path but not its timing.\n\nA line integral with respect to arc length,\n\n\\[\\int_C f\\,ds = \\int_I f(\\mathbf r(t))\\|\\mathbf r'(t)\\|\\,dt,\\]\n\nis invariant under regular one-to-one reparameterizations of the same traversal. Repeated traversal can change the integral, while orientation reversal changes the sign of a line integral of the form \\(\\int_C\\mathbf F\\cdot d\\mathbf r\\)."
    },
    {
      "heading": "Surfaces and higher-dimensional spaces",
      "text": "A surface patch in \\(\\mathbb R^3\\) is represented by\n\n\\[\\mathbf X:U\\subseteq\\mathbb R^2\\longrightarrow\\mathbb R^3, \\qquad (u,v)\\longmapsto\\mathbf X(u,v).\\]\n\nFor a smooth patch with open domain \\(U\\), the standard regularity condition is that the two partial derivatives \\(\\mathbf X_u\\) and \\(\\mathbf X_v\\) are linearly independent. Equivalently,\n\n\\[\\mathbf X_u\\times\\mathbf X_v\\ne\\mathbf0.\\]\n\nThey span the surface’s tangent space, and their cross product determines an oriented normal direction and an area-scaling factor:\n\n\\[dA=\\|\\mathbf X_u\\times\\mathbf X_v\\|\\,du\\,dv.\\]\n\nThis factor is central to a surface integral.\n\nMore generally, a smooth parameterization from an open subset of \\(\\mathbb R^k\\) into \\(\\mathbb R^n\\) has full local rank when its Jacobian matrix has rank \\(k\\). Full rank ensures independent local parameter directions, but does not by itself prevent distant parameter values from mapping to the same point.\n\nLocal coordinates and global limitations\n\nIn differential geometry, a manifold is described using compatible local coordinate charts. A chart maps a neighborhood on the manifold to an open subset of Euclidean space; its inverse is a local parameterization.\n\nA single global coordinate patch often cannot cover an entire manifold. For example, spherical coordinates give\n\n\\[\\mathbf X(\\theta,\\phi) = (\\sin\\theta\\cos\\phi,\\sin\\theta\\sin\\phi,\\cos\\theta),\\]\n\nbut the longitude \\(\\phi\\) is not unique at the poles, and the differential loses rank there. This reflects limitations of the coordinates, not a singularity of the sphere.\n\nThe sphere cannot be represented by one chart onto an open subset of \\(\\mathbb R^2\\). Multiple overlapping charts avoid this obstruction. Questions about global parameterization consequently depend on topology as well as on algebraic formulas."
    },
    {
      "heading": "Rational parameterization",
      "text": "A rational parameterization expresses coordinates as ratios of polynomials in the parameters. For example,\n\n\\[x(t)=\\frac{1-t^2}{1+t^2}, \\qquad y(t)=\\frac{2t}{1+t^2}\\]\n\nlies on the unit circle because\n\n\\[x(t)^2+y(t)^2=1.\\]\n\nFor real \\(t\\), this representation covers every point of the circle except \\((-1,0)\\), which is approached as \\(|t|\\to\\infty\\). The formulas can be derived by intersecting the circle with lines of slope \\(t\\) through that omitted point.\n\nRational parameterization is important in algebraic geometry and exact symbolic computation, but not every algebraic curve admits one. In characteristic zero, a nonsingular projective curve of genus greater than zero cannot be rationally parameterized. Thus an elliptic curve, despite having an algebraic equation, does not admit a rational parameterization of the kind available for a circle."
    },
    {
      "heading": "Parameterization in statistics",
      "text": "A parametric statistical model is a family of probability distributions\n\n\\[\\mathcal P=\\{P_\\theta:\\theta\\in\\Theta\\}.\\]\n\nThe parameter \\(\\theta\\) selects a distribution in the family.\n\nFor example, the normal distribution may be parameterized by its mean and variance:\n\n\\[\\theta=(\\mu,v),\\qquad \\mu\\in\\mathbb R,\\quad v>0,\\]\n\nwith density\n\n\\[p(x\\mid\\mu,v) = \\frac{1}{\\sqrt{2\\pi v}} \\exp\\!\\left[-\\frac{(x-\\mu)^2}{2v}\\right].\\]\n\nThe same family can instead be parameterized by mean and standard deviation, or by mean and precision \\(1/v\\).\n\nIdentification and redundancy\n\nA parameterization is identifiable if\n\n\\[P_\\theta=P_{\\theta'} \\quad\\Longrightarrow\\quad \\theta=\\theta'.\\]\n\nThis is a basic aspect of statistical identification. If different parameter values determine the same distribution, the data cannot distinguish those values solely through that model.\n\nFor example, permuting the component labels of a finite Gaussian mixture model leaves the mixture distribution unchanged. Its labeled parameterization is therefore generally non-identifiable unless equivalences or additional restrictions are taken into account.\n\nRedundancy may be convenient, but it can complicate interpretation, estimation, and uncertainty calculations.\n\nReparameterization and inference\n\nAn invertible transformation \\(\\eta=g(\\theta)\\) changes how a model is indexed without changing its distributions. Under maximum likelihood estimation, a maximizing value transforms accordingly:\n\n\\[\\widehat\\eta=g(\\widehat\\theta).\\]\n\nIn Bayesian inference, a probability distribution on parameters must also be transformed. If \\(\\theta\\) and \\(\\eta\\) are related by a smooth invertible transformation, their densities satisfy\n\n\\[p_\\eta(\\eta) = p_\\theta(g^{-1}(\\eta)) \\left|\\det Dg^{-1}(\\eta)\\right|.\\]\n\nConsequently, a prior described as uniform in one parameterization is generally not uniform in another. Omitting the Jacobian factor changes the probability model rather than merely changing its notation."
    },
    {
      "heading": "Optimization and numerical computation",
      "text": "In mathematical optimization, parameterization can enforce constraints by construction. A positive quantity can be written as\n\n\\[a=e^\\alpha,\\qquad \\alpha\\in\\mathbb R.\\]\n\nLikewise, strictly positive weights summing to one can be represented using the softmax function:\n\n\\[w_i=\\frac{e^{z_i}}{\\sum_{j=1}^{m}e^{z_j}}.\\]\n\nAdding the same constant to every \\(z_i\\) leaves the weights unchanged, so this parameterization is redundant. It also excludes exact zero weights for finite parameter values.\n\nSuch transformations replace a constrained search with an unconstrained one, but they change numerical geometry. If\n\n\\[x=F(\\theta),\\]\n\nthen\n\n\\[\\nabla_\\theta(f\\circ F) = DF(\\theta)^{\\mathsf T}\\nabla_x f(F(\\theta)).\\]\n\nOrdinary Euclidean gradient descent therefore generally follows different paths under different parameterizations, even when the represented objective values are identical.\n\nParameter scaling, nearly dependent directions, and coordinate singularities can affect convergence and numerical stability. A smooth bijective change of coordinates preserves corresponding local minima, but not necessarily convexity or algorithmic behavior.\n\nA penalty also depends on its parameterization. Applying the same-looking expression, such as a squared Euclidean norm, before and after a nonlinear change of variables generally defines different regularization."
    },
    {
      "heading": "Uses and distinctions",
      "text": "Parameterization supports geometric construction, integration, simulation, statistical modeling, and numerical solution methods. In computer-aided design, control points and basis functions parameterize curves and surfaces. In finite element methods, maps from reference elements to physical elements organize interpolation, differentiation, and integration.\n\nThe term also has a related modeling meaning: unresolved processes may be represented through formulas involving resolved quantities and adjustable coefficients. In that setting, a parameterization is an approximation scheme, not necessarily an exact coordinate description.\n\nAcross these uses, three distinctions are essential: the represented object is not its parameter values; multiple parameter values may represent the same object; and changing a parameterization may preserve the underlying object while altering coordinates, traversal, probability densities, penalties, or computational behavior."
    }
  ],
  "links": [
    "geometry",
    "statistics",
    "mathematical-function",
    "euclidean-space",
    "circle",
    "derivative",
    "chain-rule",
    "line-integral",
    "tangent-space",
    "jacobian-matrix",
    "differential-geometry",
    "manifold",
    "topology",
    "algebraic-geometry",
    "elliptic-curve",
    "probability-distribution",
    "normal-distribution",
    "statistical-identification",
    "gaussian-mixture-model",
    "maximum-likelihood-estimation",
    "bayesian-inference",
    "mathematical-optimization",
    "softmax-function",
    "gradient-descent",
    "numerical-stability",
    "regularization",
    "finite-element-method"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}