Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 1 — ∂f∗\partial f^*∂f∗ is recovered from ∂f\partial f∂f through bounded weak** nets

Proved
RockafellarMaxMono.Maximality.conj_subdiff_iff_nets

by mikedeng1 · Sep 27, 2026 · Mathlib 0df444a (Lean v4.33.1)

convex-analysisfunctional-analysismonotone-operatorsp2o-batch-p100bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Let EEE be a real Banach space with dual E∗E^*E∗ and bidual E∗∗E^{**}E∗∗, let fff be a lower semicontinuous proper convex function on EEE, and let x∗∈E∗x^* \in E^*x∗∈E∗ and x∗∗∈E∗∗x^{**} \in E^{**}x∗∗∈E∗∗. Then x∗∗∈∂f∗(x∗)x^{**} \in \partial f^*(x^*)x∗∗∈∂f∗(x∗) if and only if there exist

  1. a nonempty directed partially ordered index set III,
  2. a net {xi∗∣i∈I}\{x_i^* \mid i \in I\}{xi∗​∣i∈I} in E∗E^*E∗ converging to x∗x^*x∗ in the strong (norm) topology, and
  3. a bounded net {xi∣i∈I}\{x_i \mid i \in I\}{xi​∣i∈I} in EEE, with the same index set, converging to x∗∗x^{**}x∗∗ in the weak** topology,

such that xi∗∈∂f(xi)x_i^* \in \partial f(x_i)xi∗​∈∂f(xi​) for every i∈Ii \in Ii∈I. Here EEE is regarded as a subspace of E∗∗E^{**}E∗∗ through the canonical embedding, and the weak** topology is the weak topology induced on E∗∗E^{**}E∗∗ by E∗E^*E∗:

xi→x∗∗ weak**  ⟺  ⟨xi,y∗⟩→x∗∗(y∗)  for every y∗∈E∗.x_i \to x^{**} \text{ weak**} \iff \langle x_i, y^* \rangle \to x^{**}(y^*) \ \text{ for every } y^* \in E^* .xi​→x∗∗ weak**⟺⟨xi​,y∗⟩→x∗∗(y∗)  for every y∗∈E∗.

In a nonreflexive space ∂f∗\partial f^*∂f∗ maps E∗E^*E∗ into E∗∗E^{**}E∗∗ and is not the inverse of ∂f\partial f∂f; this proposition says that ∂f∗\partial f^*∂f∗ is nevertheless completely determined by ∂f\partial f∂f.

Formalization Note A net is a function on a type III carrying a partial order that is directed and nonempty, with convergence along the filter atTop; the index type lives in the same universe as EEE. Weak** convergence of (xi)(x_i)(xi​) to x∗∗x^{**}x∗∗ is stated as pointwise convergence on E∗E^*E∗ of the canonical images of the xix_ixi​, which is exactly convergence in the weak topology induced on E∗∗E^{**}E∗∗ by E∗E^*E∗. Boundedness is ∥xi∥≤C\|x_i\| \le C∥xi​∥≤C for a real constant CCC and all iii.

Preamble
import Mathlib
import Definitions.Def_RockafellarMaxMono_Shared_ProperConvex
import Definitions.Def_RockafellarMaxMono_Shared_Subdiff
import Definitions.Def_RockafellarMaxMono_Shared_Conj
open Filter Topology
universe u
Formal statement
namespace RockafellarMaxMono.Maximality

theorem conj_subdiff_iff_nets {E : Type u} [NormedAddCommGroup E] [NormedSpace ℝ E]
    [CompleteSpace E] (f : E → EReal) (hf : Shared.ProperConvex f) (hlsc : LowerSemicontinuous f)
    (x' : StrongDual ℝ E) (x'' : StrongDual ℝ (StrongDual ℝ E)) :
    x'' ∈ Shared.subdiff (Shared.conj f) x' ↔
      ∃ (I : Type u) (_ : PartialOrder I) (_ : IsDirected I (· ≤ ·)) (_ : Nonempty I)
        (xs' : I → StrongDual ℝ E) (xs : I → E),
        Tendsto xs' atTop (𝓝 x') ∧
        (∃ C : ℝ, ∀ i, ‖xs i‖ ≤ C) ∧
        (∀ y' : StrongDual ℝ E,
          Tendsto (fun i => NormedSpace.inclusionInDoubleDual ℝ E (xs i) y') atTop (𝓝 (x'' y'))) ∧
        ∀ i, xs' i ∈ Shared.subdiff f (xs i) := by sorry

end RockafellarMaxMono.Maximality
Source
Rockafellar, On the maximal monotonicity of subdifferential mappings, Pacific J. Math. 33 (1970), p. 211, Proposition 1
Read-back

What the Lean code literally says, in plain math · claude-opus-5-5

Setting. EEE is a real Banach space, E∗E^*E∗ is its continuous dual, and E∗∗E^{**}E∗∗ is its bidual. The canonical embedding ι:E→E∗∗\iota : E \to E^{**}ι:E→E∗∗ is given by ι(x)(y∗)=y∗(x)\iota(x)(y^*) = y^*(x)ι(x)(y∗)=y∗(x). The function f:E→R‾f : E \to \overline{\mathbb{R}}f:E→R has two hypotheses:

  • fff satisfies the imported predicate Shared.ProperConvex. It is not shown here; the name suggests "proper convex".
  • fff is lower semicontinuous for the norm topology.

Two imported operations appear, not shown here:

  • f∗:E∗→R‾f^* : E^* \to \overline{\mathbb{R}}f∗:E∗→R is Shared.conj f. The name suggests the Fenchel conjugate.
  • ∂\partial∂ is Shared.subdiff. The name suggests the subdifferential. Applied to f∗f^*f∗ at a point of E∗E^*E∗, it gives a subset of E∗∗E^{**}E∗∗. Applied to fff at a point of EEE, it gives a subset of E∗E^*E∗.

Fix x∗∈E∗x^* \in E^*x∗∈E∗ and x∗∗∈E∗∗x^{**} \in E^{**}x∗∗∈E∗∗.

Claim. x∗∗∈∂f∗(x∗)x^{**} \in \partial f^*(x^*)x∗∗∈∂f∗(x∗) holds if and only if there exist the following:

  • an index set III in the same type universe as EEE, with some partial order ≤\le≤ on III, such that III is nonempty and directed (any two indices have a common upper bound);
  • a net (xi∗)i∈I(x_i^*)_{i \in I}(xi∗​)i∈I​ in E∗E^*E∗ and a net (xi)i∈I(x_i)_{i \in I}(xi​)i∈I​ in EEE.

These must satisfy four conditions, where limits are taken along III in the direction of increasing iii:

  1. xi∗→x∗x_i^* \to x^*xi∗​→x∗ in the norm of E∗E^*E∗.
  2. The net (xi)(x_i)(xi​) is norm-bounded: there is a real CCC with ∥xi∥≤C\|x_i\| \le C∥xi​∥≤C for all iii.
  3. For every y∗∈E∗y^* \in E^*y∗∈E∗, ι(xi)(y∗)=y∗(xi)→x∗∗(y∗)\iota(x_i)(y^*) = y^*(x_i) \to x^{**}(y^*)ι(xi​)(y∗)=y∗(xi​)→x∗∗(y∗) in R\mathbb{R}R. In other words, ι(xi)→x∗∗\iota(x_i) \to x^{**}ι(xi​)→x∗∗ pointwise on E∗E^*E∗ (weak-* convergence).
  4. For every i∈Ii \in Ii∈I, xi∗∈∂f(xi)x_i^* \in \partial f(x_i)xi∗​∈∂f(xi​).

Degenerate cases. Because III must be nonempty and directed, the limits cannot hold vacuously. The bound CCC may be any real number; it is not required to be positive. Choosing III with one element turns the right-hand side into x∗∗=ι(x)x^{**} = \iota(x)x∗∗=ι(x) for some x∈Ex \in Ex∈E with x∗∈∂f(x)x^* \in \partial f(x)x∗∈∂f(x). If E={0}E = \{0\}E={0}, all three spaces are {0}\{0\}{0} and both sides concern only 0∈∂f∗(0)0 \in \partial f^*(0)0∈∂f∗(0) and 0∈∂f(0)0 \in \partial f(0)0∈∂f(0).

Human review
  • Endorsed by Shuze Chen · Sep 27, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Sep 27, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me