Canonical: https://knowlpedia.com/real-analysis/moment-matching/

# Matching prescribed integral moments

Choosing a function so a specified finite family of integral functionals takes prescribed values.

Given integral functionals \(\Phi_1,\ldots,\Phi_m\) on a class of functions, **moment matching** asks for \(u\) such that
\[
\Phi_j(u)=d_j,\qquad 1\le j\le m.
\]
The **moment map** in this sense is \(\Phi(u)=(\Phi_1(u),\ldots,\Phi_m(u))\). The functionals may be linear [weighted integrals](https://knowlpedia.com/real-analysis/weighted-radial-moment/) or integrals of nonlinear expressions such as \(u^2\).

## Function-valued targets

For profiles \(u(r,\eta)\), one may prescribe \(\Phi_j(u)(\eta)=d_j(\eta)\) for every \(\eta\). There are finitely many rows, but each row is a function. Solving at finitely many sample values does not establish this identity. Corrections built from fixed radial bumps and parameter-dependent coefficients reduce the problem to a parameterized finite system. This integral moment map is distinct from a Hamiltonian moment map in symplectic geometry.
