Stage 2 closure: below-two geometry #
This file combines the native conjugate-smoothness Hessian bound with the
explicit duality/Fenchel reduction. The exported theorem is the frozen
O3.BelowGeometryStatement without additional hypotheses.
theorem
O3.Stage2Closure.squaredLpEnergy_two_smooth
{d : ℕ}
(u v : Point d)
:
Stage2RouteB.squaredLpEnergy 2 v ≤ Stage2RouteB.squaredLpEnergy 2 u + pairing (dualityMap 2 u) (v - u) + (2 - 1) / 2 * lpNorm 2 (v - u) ^ 2
theorem
O3.Stage2Closure.hasDerivAt_squaredLpEnergy_line
{q : ℝ}
(hq : 1 < q)
{d : ℕ}
(u h : Point d)
(t : ℝ)
:
HasDerivAt (fun (s : ℝ) => Stage2RouteB.squaredLpEnergy q (u + s • h)) (pairing (dualityMap q (u + t • h)) h) t
The squared norm along a line after subtracting its smoothness quadratic.
Equations
Instances For
The directional derivative of the smoothness remainder along a line.
Equations
- O3.Stage2Closure.smoothRemainderDeriv q u h t = O3.pairing (O3.dualityMap q (u + t • h)) h - (q - 1) * t * O3.lpNorm q h ^ 2
Instances For
theorem
O3.Stage2Closure.hasDerivAt_smoothRemainder
{q : ℝ}
(hq : 1 < q)
{d : ℕ}
(u h : Point d)
(t : ℝ)
:
HasDerivAt (smoothRemainder q u h) (smoothRemainderDeriv q u h t) t
theorem
O3.Stage2Closure.exists_hasDerivAt_smoothRemainderDeriv_nonpos
{q : ℝ}
(hq : 2 < q)
{d : ℕ}
(u h : Point d)
(t : ℝ)
:
∃ (w : ℝ), HasDerivAt (smoothRemainderDeriv q u h) w t ∧ w ≤ 0
theorem
O3.Stage2Closure.smoothRemainder_concave_above_two
{q : ℝ}
(hq : 2 < q)
{d : ℕ}
(u h : Point d)
:
ConcaveOn ℝ Set.univ (smoothRemainder q u h)
theorem
O3.Stage2Closure.squaredLpEnergy_smooth_above_two
{q : ℝ}
(hq : 2 < q)
{d : ℕ}
(u v : Point d)
:
Stage2RouteB.squaredLpEnergy q v ≤ Stage2RouteB.squaredLpEnergy q u + pairing (dualityMap q u) (v - u) + (q - 1) / 2 * lpNorm q (v - u) ^ 2
Frozen Stage 2 target: exact strong convexity of the squared finite
ell_p norm for every real 1 < p ≤ 2 and every finite dimension.