Skip to content

Secondary Measure

An auxiliary positive measure derived from an initial moment measure so that its secondary-polynomial sequence becomes orthogonal, with the pair linked by a reciprocal Stieltjes-transform relation.

Version
v3 · 2026-09-06 · History
Domain-specific #
2733
Origin domain
mathematics
Subdomain
orthogonal polynomials
Aliases
Secondary orthogonality measure, Associated secondary measure

Core Idea

Let a positive measure \(\rho\) with sufficiently many finite moments generate orthogonal polynomials \(P_n\). Their associated or secondary polynomials \(Q_n\) can be defined by a divided-difference transform such as

\[ Q_n(x)=\int \frac{P_n(t)-P_n(x)}{t-x}\,d\rho(t), \]

with convention-dependent signs and normalizations. A secondary measure \(\mu\) associated with \(\rho\) is a positive measure for which the \(Q_n\) form an orthogonal family.[1]

For a normalized initial measure and a declared Stieltjes-transform convention, the construction is encoded by a reciprocal relation of the form

\[ S_\mu(z)=z-c_1-\frac{1}{S_\rho(z)}, \]

up to a positive scaling constant, where \(c_1\) is the first moment of \(\rho\).[2] The recognition invariant is initial moment measure + secondary-polynomial operator + positive orthogonalizing measure + reciprocal Stieltjes relation.

Structural Signature

  • A positive initial measure on a declared real interval or support.
  • Finite moments to the order required, often of every order.
  • An orthogonal polynomial family \(P_n\) fixed by a normalization.
  • Secondary polynomials \(Q_n\) obtained from a divided-difference operator.
  • A candidate positive measure \(\mu\).
  • Orthogonality of the \(Q_n\) in \(L^2(\mu)\).
  • A Stieltjes or Cauchy transform for each measure.
  • A reciprocal transform relation fixing \(\mu\) up to stated scale.
  • Existence and positivity conditions, not merely a formal series identity.
  • Moment normalization, including the mass of \(\mu\).
  • Norm correspondence between \(P_n\) and \(Q_n\) under the standard normalization.
  • Optional iteration by normalizing \(\mu\) and repeating the construction.

What It Is Not

It is not any measure mentioned after another measure, nor a conditional, pushforward, or quasi-invariant measure. Its defining work is to orthogonalize the secondary polynomial sequence attached to the initial orthogonal system.

It is not guaranteed to exist as a positive density for arbitrary input data. A formal transform or moment sequence must satisfy the analytic and positivity conditions needed to represent a measure.[3]

Scope of Application

Secondary measures belong to the theory of orthogonal polynomials, Jacobi recurrences, moment problems, and Stieltjes transforms. They support operator identities between weighted \(L^2\) spaces, explicit calculations for classical polynomial families, and iterative sequences of normalized measures. Examples associated with shifted Legendre, Laguerre, Hermite, and Chebyshev systems illustrate how the transform relation can produce explicit densities and Fourier coefficients.[1]

Clarity

State the sign convention for \(S_\rho\), whether measures are normalized to probability mass, the normalization of \(P_n\) and \(Q_n\), the interval, and all moment or density hypotheses. Without these declarations, formulas that differ only conventionally may look contradictory.

Manages Complexity

The construction replaces a direct search for an orthogonalizing weight for \(Q_n\) with a transform-level update. Recurrence coefficients, moments, and analytic boundary values can then be used systematically instead of recomputing every pairwise integral.

Abstract Reasoning

  1. Fix \(\rho\), its support, moments, and Stieltjes-transform convention.
  2. Construct and normalize the orthogonal polynomials \(P_n\).
  3. Apply the divided-difference operator to obtain \(Q_n\).
  4. Derive the reciprocal Stieltjes expression proposed for \(\mu\).
  5. Verify that the expression is the transform of a positive measure.
  6. Recover its density or moments when possible.
  7. Check orthogonality and norm relations for the \(Q_n\).
  8. Normalize and iterate only after recording the mass introduced at each step.

Knowledge Transfer

The portable pattern is to construct a new carrier that makes a transformed basis orthogonal, using an analytic transform to move between carriers. The proposed immediate parent is Measure.

Examples

Lebesgue weight on \([0,1]\). The shifted Legendre system has a secondary density expressible through the logarithmic boundary-value reducer.

Normalized iteration. Dividing a secondary measure by its total mass produces a new probability measure to which the construction can be applied again, yielding a sequence tied to shifted recurrence data.

Fixed point behavior. Under a particular normalization, a Chebyshev-type measure can reproduce its own secondary measure, illustrating that the operation is not merely an arbitrary reweighting.

Structural Tensions

  • Formal transform identity versus positive-measure existence.
  • Canonical construction versus scaling freedom.
  • Explicit density versus moment-only specification.
  • Orthogonal-polynomial normalization versus invariant content.
  • Iterative elegance versus accumulating convention errors.
  • Reducible cases versus measures for which operator inversion fails.

Structural–Framed Character

Transform, derived carrier, and orthogonalization are structural. Positive measures, moments, orthogonal polynomials, and Stieltjes boundary values are mathematical frame.

Structural Core vs. Domain Accent

The portable core is changing the measure so a transformed family acquires orthogonality. The constitutive accent is the precise secondary-polynomial operator and reciprocal Stieltjes-transform relation.

Measure is the proposed immediate parent. Transformation, Orthogonality, Duality, Normalization, Recurrence, and Fixed Point are related.

The prospective queue contains one strict edge to prime:measure. No live DAG mutation is authorized.

Relationships to Other Abstractions

Local relationship map for Secondary MeasureParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Secondary MeasureDOMAINPrime abstraction: Measure — is a kind ofMeasurePRIME

Current abstraction Secondary Measure Domain-specific

Parents (1) — more general patterns this builds on

  • Secondary Measure is a kind of Measure Prime

    Measure is the proposed immediate parent.

Hierarchy paths (2) — routes to 2 parentless roots

Neighborhood in Abstraction Space

Secondary Measure sits in a sparse region of the domain-specific corpus (89th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Moment Problems & Discrete Approximation (7 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With

  • Quasi-invariant measure.
  • Pushforward or image measure.
  • Conditional probability measure.
  • Spectral measure without the secondary-polynomial relation.
  • Associated polynomials without an orthogonalizing measure.
  • A formal Stieltjes series not representing a positive measure.

References

[1] Roland Groux, “About the Operator Creating Secondary Polynomials,” arXiv:1104.3218 (2011), https://arxiv.org/abs/1104.3218. registry ↩a ↩b

[2] Roland Groux, “Some Explicit Formulas for a Sequence of Secondary Measures,” arXiv:1104.4559 (2011), https://arxiv.org/abs/1104.4559. registry ↩

[3] Theodore S. Chihara, An Introduction to Orthogonal Polynomials (Gordon and Breach, 1978), especially the moment-functional and three-term-recurrence framework. registry ↩

[4] Barry Simon, Orthogonal Polynomials on the Real Line, AMS Colloquium Publications 54, Part 1 (2005), for Jacobi parameters, spectral measures, and Stieltjes transforms. registry ↩