Limits of positive-definite functions #
This file derives pointwise- and locally-uniform-limit closure for
TauCeti.IsPositiveDefinite, the positive-definite function predicate on an involutive additive
monoid, from Mathlib's closedness theorem for the cone of positive-semidefinite matrices.
This is the limit-closure item from Part C of the OneParameterSemigroups roadmap. The result is
about positive-definiteness alone for pointwise limits; as the roadmap notes, continuity needs an
additional hypothesis. Mathlib's TendstoLocallyUniformly.continuous provides continuity of a
locally uniform limit when the approximating functions are frequently continuous, while this file
provides the positive-definiteness conclusion.
Main declarations #
TauCeti.IsPositiveDefinite.of_tendsto: filter-level pointwise-limit closure for positive-definite functions.TauCeti.IsPositiveDefinite.of_forall_tendsto: the same result when every function in the family is positive definite.TauCeti.IsPositiveDefinite.of_seq_tendsto: the sequentialatTopspecialization.TauCeti.IsPositiveDefinite.of_tendstoLocallyUniformly: locally uniform limits of eventually positive-definite functions are positive definite.
References #
- C. Berg, J. P. R. Christensen, P. Ressel, Harmonic Analysis on Semigroups (GTM 100, 1984), Chapter 3.
Positive-definiteness is preserved under pointwise limits along a nontrivial filter. The hypothesis on positive-definiteness is eventual, so this applies equally to nets that are eventually positive definite.
A pointwise limit of a family of positive-definite functions is positive definite. This is the
non-eventual form of TauCeti.IsPositiveDefinite.of_tendsto.
Sequential pointwise limits of eventually positive-definite functions are positive definite.
A locally uniform limit of eventually positive-definite functions is positive definite.
Unlike continuity, positive-definiteness itself only needs pointwise convergence; local uniform
convergence is converted to pointwise convergence before applying
TauCeti.IsPositiveDefinite.of_tendsto.