{
  "id": "polynomial-interpolation",
  "title": "Polynomial Interpolation",
  "lang": "en",
  "source_lang": "en",
  "summary": "Polynomial interpolation constructs a polynomial that matches prescribed values at distinct nodes, providing a fundamental tool for numerical approximation, computation, and algebra.",
  "category": "mathematics",
  "subcategory": "applied",
  "type": "concept",
  "aliases": [
    "Interpolating polynomial construction"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/polynomial-interpolation/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/polynomial-interpolation/",
      "json": "https://aiwiki.page/zh/polynomial-interpolation.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Polynomial interpolation is the construction of a polynomial that takes specified values at a finite set of distinct input points, called nodes. Given \\(n+1\\) distinct nodes and their associated values, there is exactly one polynomial of degree at most \\(n\\) satisfying all the prescribed conditions. It is a central technique in interpolation, numerical analysis, and approximation theory, with applications in numerical integration, differentiation, differential equations, and algebraic computation.\n\nExact agreement at the nodes does not necessarily imply good approximation between them. The placement of the nodes, the behavior of the underlying function, and the numerical method used to represent and evaluate the polynomial all influence the quality of the result."
    },
    {
      "heading": "Definition and uniqueness",
      "text": "Let \\(x_0,\\ldots,x_n\\) be distinct real numbers, and let \\(y_0,\\ldots,y_n\\) be prescribed values. The interpolation problem is to find\n\n\\[p_n(x)=a_0+a_1x+\\cdots+a_nx^n\\]\n\nsuch that\n\n\\[p_n(x_i)=y_i,\\qquad i=0,\\ldots,n.\\]\n\nThe degree bound is essential: without it, infinitely many polynomials satisfy the conditions. Once one solution \\(p_n\\) is known, every polynomial of the form\n\n\\[p_n(x)+q(x)\\prod_{i=0}^{n}(x-x_i),\\]\n\nwhere \\(q\\) is any polynomial, has the same values at the nodes.\n\nUniqueness follows from the root bound for polynomials. If \\(p_n\\) and \\(r_n\\) both solve the problem, then \\(p_n-r_n\\) has degree at most \\(n\\) but has \\(n+1\\) distinct roots. It must therefore be the zero polynomial.\n\nExistence follows, for example, from the explicit Lagrange formula. More abstractly, evaluation at the nodes is an invertible linear map from the \\((n+1)\\)-dimensional vector space of polynomials of degree at most \\(n\\) to the space of data vectors.\n\nThe same existence and uniqueness statement holds over any field, including the complex numbers and finite fields, provided the nodes are distinct elements of that field."
    },
    {
      "heading": "Principal representations",
      "text": "Different interpolation formulas describe the same polynomial. Their differences concern how it is constructed, updated, stored, and evaluated.\n\nMonomial form and the Vandermonde system\n\nSubstituting the interpolation conditions into the monomial expansion gives a system of linear equations:\n\n\\[\\begin{pmatrix} 1&x_0&x_0^2&\\cdots&x_0^n\\\\ 1&x_1&x_1^2&\\cdots&x_1^n\\\\ \\vdots&\\vdots&\\vdots&&\\vdots\\\\ 1&x_n&x_n^2&\\cdots&x_n^n \\end{pmatrix} \\begin{pmatrix} a_0\\\\a_1\\\\\\vdots\\\\a_n \\end{pmatrix} = \\begin{pmatrix} y_0\\\\y_1\\\\\\vdots\\\\y_n \\end{pmatrix}.\\]\n\nThe coefficient matrix is a Vandermonde matrix. Its determinant is\n\n\\[\\prod_{0\\le i<j\\le n}(x_j-x_i),\\]\n\nwhich is nonzero precisely when the nodes are distinct.\n\nThis representation is conceptually direct, but the monomial coefficients can be highly sensitive to small perturbations. Solving a Vandermonde system and then evaluating the resulting monomial expansion is therefore not generally the preferred way to perform interpolation.\n\nLagrange form\n\nDefine the cardinal polynomials\n\n\\[\\ell_i(x)= \\prod_{\\substack{0\\le j\\le n\\\\j\\ne i}} \\frac{x-x_j}{x_i-x_j}.\\]\n\nThey satisfy\n\n\\[\\ell_i(x_j)= \\begin{cases} 1,&i=j,\\\\ 0,&i\\ne j. \\end{cases}\\]\n\nThe interpolating polynomial is\n\n\\[p_n(x)=\\sum_{i=0}^{n}y_i\\ell_i(x).\\]\n\nThis formula proves existence directly and separates the role of the nodes from the role of the data. The cardinal polynomials depend only on the nodes; changing the values changes only their coefficients.\n\nIt also shows that interpolation is linear in the data and reproduces every polynomial of degree at most \\(n\\) exactly.\n\nNewton form and divided differences\n\nThe Newton representation is\n\n\\[p_n(x)=c_0+c_1(x-x_0) +c_2(x-x_0)(x-x_1) +\\cdots+ c_n\\prod_{j=0}^{n-1}(x-x_j).\\]\n\nIts coefficients are divided differences. For data sampled from a function \\(f\\), they are defined recursively by\n\n\\[f[x_i]=f(x_i)\\]\n\nand\n\n\\[f[x_i,\\ldots,x_{i+k}] = \\frac{ f[x_{i+1},\\ldots,x_{i+k}] -f[x_i,\\ldots,x_{i+k-1}] }{ x_{i+k}-x_i }.\\]\n\nThen\n\n\\[c_k=f[x_0,\\ldots,x_k].\\]\n\nA standard divided-difference table requires \\(O(n^2)\\) arithmetic operations, while nested evaluation of the resulting polynomial requires \\(O(n)\\) operations per point. Newton form is particularly convenient when nodes are added: the existing expansion is retained and one further term is appended.\n\nBarycentric form\n\nA widely used evaluation formula is barycentric interpolation. Define the weights\n\n\\[w_i= \\frac{1}{ \\displaystyle\\prod_{\\substack{0\\le j\\le n\\\\j\\ne i}}(x_i-x_j) }.\\]\n\nFor \\(x\\) distinct from every node,\n\n\\[p_n(x)= \\frac{ \\displaystyle\\sum_{i=0}^{n}\\frac{w_i y_i}{x-x_i} }{ \\displaystyle\\sum_{i=0}^{n}\\frac{w_i}{x-x_i} }.\\]\n\nAlthough this is written as a quotient, in exact arithmetic it equals the polynomial interpolant. At a node, the value is assigned directly as \\(p_n(x_i)=y_i\\).\n\nOnce the weights are available, evaluation costs \\(O(n)\\) operations. Multiplying all weights by the same nonzero constant leaves the formula unchanged, allowing convenient rescaling. For several important node families, explicit weights avoid the direct product computation.\n\nBarycentric evaluation often has favorable numerical stability, especially for well-chosen nodes within the interpolation interval. It does not, however, eliminate sensitivity inherent in the interpolation problem itself."
    },
    {
      "heading": "Example",
      "text": "Consider the three data points\n\n\\[(0,1),\\qquad(1,3),\\qquad(2,2).\\]\n\nTheir divided differences are\n\n\\[f[0,1]=2,\\qquad f[1,2]=-1,\\qquad f[0,1,2]=-\\frac32.\\]\n\nThe Newton interpolant is therefore\n\n\\[p_2(x)=1+2x-\\frac32x(x-1) =1+\\frac72x-\\frac32x^2.\\]\n\nIt takes the prescribed values at all three nodes. At the intermediate point \\(x=\\tfrac12\\), it gives\n\n\\[p_2\\!\\left(\\frac12\\right)=\\frac{19}{8}.\\]\n\nThis is the value of the interpolating polynomial, not necessarily the value of an unknown function that generated the observations."
    },
    {
      "heading": "Approximation error",
      "text": "Suppose the data are exact samples \\(y_i=f(x_i)\\), where \\(f\\) has a continuous \\((n+1)\\)st derivative on an interval containing the nodes and the evaluation point \\(x\\). For \\(x\\) not equal to a node, there exists a point \\(\\xi\\) in the smallest interval containing these points such that\n\n\\[f(x)-p_n(x)= \\frac{f^{(n+1)}(\\xi)}{(n+1)!} \\prod_{i=0}^{n}(x-x_i).\\]\n\nAt a node, the error is zero. Consequently, on an interval \\([a,b]\\) containing all nodes,\n\n\\[|f(x)-p_n(x)| \\le \\frac{M_{n+1}}{(n+1)!} \\left|\\prod_{i=0}^{n}(x-x_i)\\right|,\\]\n\nwhere\n\n\\[M_{n+1}=\\max_{t\\in[a,b]}|f^{(n+1)}(t)|.\\]\n\nThe formula separates two influences: the smoothness and derivative growth of the function, and the geometry of the nodes through the nodal product.\n\nIt does not imply that increasing the degree always improves the approximation. The derivative bound itself depends on \\(n\\), and the nodal product can behave poorly. The formula also concerns exact function values, not errors introduced by measurement noise or finite-precision computation."
    },
    {
      "heading": "Node selection and convergence",
      "text": "Runge phenomenon\n\nHigh-degree interpolation at equally spaced nodes can exhibit large oscillations near the ends of an interval. The classical Runge phenomenon occurs for\n\n\\[f(x)=\\frac{1}{1+25x^2}, \\qquad -1\\le x\\le 1.\\]\n\nAs the number of equally spaced nodes increases, the interpolants fail to converge uniformly to this smooth function.\n\nThis behavior does not contradict polynomial approximation theorems: the existence of good polynomial approximations does not mean that interpolating at arbitrary prescribed nodes produces them.\n\nChebyshev nodes\n\nA standard alternative on \\([-1,1]\\) is the set of Chebyshev nodes\n\n\\[x_i=\\cos\\!\\left(\\frac{(2i+1)\\pi}{2(n+1)}\\right), \\qquad i=0,\\ldots,n.\\]\n\nThese nodes cluster near the endpoints. They are the roots of the Chebyshev polynomial \\(T_{n+1}\\), a member of a classical family of orthogonal polynomials.\n\nAmong all monic polynomials of degree \\(n+1\\), the nodal polynomial associated with these roots minimizes the maximum absolute value on \\([-1,1]\\):\n\n\\[\\max_{x\\in[-1,1]} \\left|\\prod_{i=0}^{n}(x-x_i)\\right| =2^{-n}.\\]\n\nA related choice, the Chebyshev–Lobatto nodes, includes both endpoints:\n\n\\[x_i=\\cos\\!\\left(\\frac{i\\pi}{n}\\right), \\qquad i=0,\\ldots,n,\\quad n\\ge1.\\]\n\nBoth families substantially improve high-degree interpolation behavior relative to equally spaced nodes. If a function extends analytically to a suitable complex neighborhood of the interval, Chebyshev interpolation converges geometrically. For functions of limited smoothness, convergence is generally slower.\n\nNo fixed sequence of node sets guarantees uniform convergence of the polynomial interpolants for every continuous function on a nondegenerate compact interval."
    },
    {
      "heading": "Conditioning, stability, and data errors",
      "text": "The sensitivity of interpolation to perturbations in the sampled values is described by the Lebesgue function\n\n\\[\\Lambda_n(x)=\\sum_{i=0}^{n}|\\ell_i(x)|\\]\n\nand the Lebesgue constant\n\n\\[\\Lambda_n=\\max_{x\\in[a,b]}\\Lambda_n(x).\\]\n\nIf the data are perturbed by amounts satisfying \\(|\\delta y_i|\\le\\varepsilon\\), then the change in the exact interpolant satisfies\n\n\\[|\\delta p_n(x)|\\le\\Lambda_n(x)\\varepsilon,\\]\n\nand hence\n\n\\[\\|\\delta p_n\\|_\\infty\\le\\Lambda_n\\varepsilon.\\]\n\nThus the Lebesgue constant acts as an absolute condition number for the map from data values to the interpolating function in the maximum norm.\n\nIt also relates interpolation to best approximation. If\n\n\\[E_n(f)=\\inf_{\\deg q\\le n}\\|f-q\\|_\\infty,\\]\n\nthen\n\n\\[\\|f-p_n\\|_\\infty \\le(1+\\Lambda_n)E_n(f).\\]\n\nFor Chebyshev nodes, \\(\\Lambda_n\\) grows only logarithmically with \\(n\\); for equally spaced nodes, it grows exponentially.\n\nConditioning must be distinguished from algorithmic stability. A stable implementation can limit additional error from floating-point arithmetic, but cannot remove amplification inherent in the selected nodes. Conversely, monomial coefficient recovery may be badly conditioned even when values of the interpolant can be evaluated accurately in a different representation.\n\nEvaluation outside the span of the nodes is extrapolation. There, the cardinal polynomials and the associated sensitivity can grow rapidly."
    },
    {
      "heading": "Related methods and extensions",
      "text": "Piecewise polynomial interpolation. Instead of fitting one polynomial over an entire interval, a method can fit low-degree polynomials on subintervals. Spline interpolation imposes continuity conditions where the pieces meet. It often avoids the endpoint oscillations of high-degree global interpolation, although the locality of data influence depends on the spline construction.\n\nHermite interpolation. Hermite interpolation prescribes derivative values as well as function values. In the standard univariate construction, a total of \\(m\\) conditions consisting of consecutive derivatives at distinct nodes determines a unique polynomial of degree at most \\(m-1\\). Interpolation of a function and its first \\(n\\) derivatives at a single point gives its degree-\\(n\\) Taylor polynomial.\n\nLeast-squares approximation. Interpolation enforces exact agreement at the data points. Least-squares approximation instead minimizes an aggregate residual and need not pass through every observation. With noisy data, exact interpolation may reproduce unwanted fluctuations; smoothing and regularization modify the fitting problem rather than solving the same interpolation conditions.\n\nMultivariate interpolation. In several variables, the number of points alone does not guarantee uniqueness: their geometric arrangement matters. For example, three points determine a unique affine polynomial in two variables only when they are not collinear. Tensor-product grids and interpolation on simplices provide structured constructions."
    },
    {
      "heading": "Applications",
      "text": "Polynomial interpolation supplies a common foundation for several computational methods:\n\n- Numerical integration: integrating the Lagrange representation gives quadrature formulas. Equally spaced nodes lead to Newton–Cotes formulas; specially selected nodes yield other families with different accuracy and stability properties.\n- Numerical differentiation: differentiating an interpolating polynomial produces derivative approximations from sampled values and provides a systematic derivation of finite-difference weights.\n- Differential equations: collocation and spectral methods represent approximate solutions using polynomials and impose governing equations at selected points.\n- Finite elements: polynomial shape functions in the finite element method interpolate local degrees of freedom on individual mesh elements.\n- Algebraic reconstruction: interpolation over finite fields reconstructs polynomials from exact samples and underlies constructions used in error-correcting codes and secret sharing.\n\nIn each setting, polynomial degree, node placement, data accuracy, and representation determine whether interpolation serves as a reliable computational building block."
    }
  ],
  "links": [
    "polynomial",
    "interpolation",
    "numerical-analysis",
    "linear-map",
    "vector-space",
    "field-mathematics",
    "finite-field",
    "system-of-linear-equations",
    "determinant",
    "numerical-stability",
    "derivative",
    "runge-phenomenon",
    "orthogonal-polynomials",
    "uniform-convergence",
    "continuous-function",
    "condition-number",
    "floating-point-arithmetic",
    "taylor-series",
    "least-squares",
    "regularization",
    "finite-element-method"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}