openAI-math-dataset / lean /OAI /NumberTheory /Jacobsthal /Partitions /GeometricTotalBoxMass.lean
roseking's picture
Add lean source and binary files (part 104)
f89a930 verified
Raw History Blame Contribute Delete
3.17 kB
import OAI.NumberTheory.Jacobsthal.Paths.OutsideExceptionsWitness
namespace OAI
namespace Erdos970
open scoped _root_.Erdos970
namespace NumberTheoryLean.GeometricTotalBoxMass
open FinitePathGeometry PrimeHistories PrimeBinMembership ActualSafeBoxMass
open GeometricBoxImages SafeSubsetBoxGeometry CanonicalSubsetBox FilteredBoxWordMass
open LogarithmicBinScale ErdosSubsetWord
theorem uniform_geometric_box_mass (K : ℝ) (hK : 3 ≤ K) :
∃ C B₀ w₀ : ℝ,0 < C ∧ 3 ≤ B₀ ∧ 1 < w₀ ∧
∀ B w top : ℝ,B₀ ≤ B → w₀ ≤ w → ∀ (hw : 1 < w) (htop : w < top),
∀ xi Clen : ℝ,∀ hxi : 0 < xi,1 ≤ Clen → Real.log B ≤ 2*Real.log w → 2*Clen*xi ≤ 1 →
∀ z : Node,z.side=.even → 199/100 ≤ z.ratio → z.ratio ≤ 23/10 → Consistent z →
z.cutoff=B → z.closed=true → w^B=top → ∀ L alpha beta : ℝ,
B^2*(∑ m∈geometricBoxes hw htop hxi Clen B K L alpha beta z,selectionMass (globalBins w top xi) m) ≤ C := by
obtain ⟨C,B₀,W,hC,hB₀,hW,hTotal⟩ := uniform_safe_box_mass K hK
refine ⟨C,B₀,max W (Real.exp 1),hC,hB₀,hW.trans_le (le_max_left _ _),?_⟩
intro B w top hB hw₀ hw htop xi Clen hxi hClen hcomp hsmall z hi h199 h23 hz hcut hclosed hpower L alpha beta
have hlog : 1 ≤ Real.log w := by
have hh := Real.log_le_log (Real.exp_pos 1) ((le_max_right _ _).trans hw₀)
simpa only [Real.log_exp] using hh
have hmesh : 2*(xi/Real.log w) ≤ 6*(2*Clen*xi) := by
have hh := div_le_self hxi.le hlog
nlinarith
exact hTotal B w top hB ((le_max_left _ _).trans hw₀) hw htop xi Clen hxi
(by linarith) hcomp hsmall z hi h199 h23 hz hcut hclosed hpower
(geometricBoxes hw htop hxi Clen B K L alpha beta z)
(geometric_box_anchor hw htop hxi (by linarith) hcomp hmesh z)
theorem unlisted_box_transfer {w top xi Cs C B K L alpha beta sigma A : ℝ}
(hw : 1 < w) (htop : w < top) (hxi : 0 < xi) (hsigma : 0 ≤ sigma)
(a : ℕ → ℕ) (z : Node) (Q : (Fin (binCount w top xi) → ℕ) → Finset ℚ)
(F : Finset (List ℕ))
(hTotal : B^2*(∑ m∈geometricBoxes hw htop hxi C B K L alpha beta z,selectionMass (globalBins w top xi) m) ≤ A)
(hBox : ∀ m∈geometricBoxes hw htop hxi C B K L alpha beta z,
selectedWordMass (globalBins w top xi) m
(ExpandedCompactPartition.unlistedGeometric hw htop hxi Cs C B K L alpha beta a z Q F)
≤ sigma*selectionMass (globalBins w top xi) m) :
B^2*(∑ ps∈ExpandedCompactPartition.unlistedGeometric hw htop hxi Cs C B K L alpha beta a z Q F,
ErdosPrimeInputs.PrimePrefixMass.prefixWeight ps) ≤ sigma*A := by
rw [unlisted_geometric_box_mass hw htop hxi a z Q F]
calc
_ ≤ B^2*(∑ m∈geometricBoxes hw htop hxi C B K L alpha beta z,sigma*selectionMass (globalBins w top xi) m) :=
mul_le_mul_of_nonneg_left (Finset.sum_le_sum hBox) (sq_nonneg B)
_ = sigma*(B^2*(∑ m∈geometricBoxes hw htop hxi C B K L alpha beta z,selectionMass (globalBins w top xi) m)) := by
rw [← Finset.mul_sum]
ring
_ ≤ _ := mul_le_mul_of_nonneg_left hTotal hsigma
end NumberTheoryLean.GeometricTotalBoxMass
end Erdos970
end OAI