parallel_mw_balancing_numerator_bound
Under the displayed size and power inequalities, the triple numerator is at least B times the balanced-endpoint benchmark divided by two.
Statement
theorem parallel_mw_balancing_numerator_bound (a c A B C : ℝ)
(ha : 1 ≤ a) (hc : 1 ≤ c) (hA : a + 1 ≤ A) (hC : c + 1 ≤ C)
(hquad : a^2 + a ≤ 2*A) (hgap : a = 1 ∨ 2 ≤ a)
(hsmall : a = 1 → A = 2) (hB : 2*A ≤ B) :
B*(a*(a+2)+2*(a+1)*c) ≤ 2*(2*A*B*C-2*A-4*B-2*C+6)- Source
- Algebraic reduction for parallel_mw_ratio_decreasing_triple and OP-670.
- Verified
- 02 Nov 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 1
- parallel_mw_balancing_base_polynomialA polynomial lower bound under natural-number-style lower bounds on real a and c and exponential proxies A and C.
- Used by 1
- parallel_mw_ratio_decreasing_tripleFor positive integers a, b, c with a + 2 ≤ b, replacing (a, b) with (a + 1, b - 1) strictly decreases the rational formula for…
The proof, explained
The proof views the target inequality as a linear function of B for B ≥ 2*A. It first invokes parallel_mw_balancing_base_polynomial to show the inequality holds at the benchmark value B = 2*A. It then proves that the slope in B, namely 4*A*C - 8 - (a*(a+2) + 2*(a+1)*c), is non-negative by combining the base polynomial with the lower bound A + C ≥ 3 and dividing out A > 0. Finally, non-linear arithmetic completes the proof by expressing the difference between the two sides as the sum of twice the base polynomial and the non-negative slope scaled by B - 2*A.
0 ≤ 4*A^2*C - 10*A - 2*C + 6 - A*(a*(a+2) + 2*(a+1)*c)
Applying the previously proven lemma parallel_mw_balancing_base_polynomial under the given hypotheses establishes the bound for the base case where B = 2*A.
uses
parallel_mw_balancing_base_polynomialhave hp := parallel_mw_balancing_base_polynomial a c A C ha hc hA hC hquad hgap hsmall0 < A and 0 ≤ A + C - 3
From 1 ≤ a and a + 1 ≤ A, we obtain A ≥ 2 > 0. Similarly, 1 ≤ c and c + 1 ≤ C yield C ≥ 2, so A + C - 3 ≥ 2 + 2 - 3 = 1 ≥ 0. Linear arithmetic establishes both inequalities.
have hApos : 0 < A := by linarith have hAC : 0 ≤ A+C-3 := by linarith
0 ≤ A * (4*A*C - 8 - (a*(a+2) + 2*(a+1)*c))
The polynomial expression expands to the base polynomial plus 2*(A + C - 3). Because the base polynomial and A + C - 3 are both non-negative, non-linear arithmetic verifies that this product is non-negative.
have hAk : 0 ≤ A*(4*A*C-8-(a*(a+2)+2*(a+1)*c)) := by nlinarith
0 ≤ 4*A*C - 8 - (a*(a+2) + 2*(a+1)*c)
Suppose for contradiction that this factor is strictly negative. Since A > 0, mul_neg_of_pos_of_neg implies that multiplying it by A yields a strictly negative product, contradicting the non-negativity proven in the previous step.
uses
mul_neg_of_pos_of_neghave hk : 0 ≤ 4*A*C-8-(a*(a+2)+2*(a+1)*c) := by by_contra hh have hn : 4*A*C-8-(a*(a+2)+2*(a+1)*c) < 0 := by linarith have hh' := mul_neg_of_pos_of_neg hApos hn linarith
0 ≤ (B - 2*A) * (4*A*C - 8 - (a*(a+2) + 2*(a+1)*c))
Since 2*A ≤ B, the factor B - 2*A is non-negative by linear arithmetic. Applying mul_nonneg to this factor and the non-negative slope from the previous step shows that their product is non-negative.
uses
mul_nonneghave hm := mul_nonneg (show 0 ≤ B-2*A by linarith) hk
B * (a*(a+2) + 2*(a+1)*c) ≤ 2 * (2*A*B*C - 2*A - 4*B - 2*C + 6)
The difference between the right-hand side and the left-hand side expands algebraically to twice the base polynomial plus (B - 2*A) * (4*A*C - 8 - (a*(a+2) + 2*(a+1)*c)). Since both components have been shown to be non-negative, non-linear arithmetic closes the goal.
nlinarith
parallel_mw_balancing_base_polynomial- For real numbers satisfying 1 ≤ a, 1 ≤ c, a + 1 ≤ A, c + 1 ≤ C, a² + a ≤ 2A, (a = 1 ∨ 2 ≤ a), and (a = 1 → A = 2), the polynomial 4A²C - 10A - 2C + 6 - A(a(a + 2) + 2(a + 1)c) is non-negative.
mul_neg_of_pos_of_neg- The product of a strictly positive number and a strictly negative number is strictly negative (0 < a → b < 0 → a * b < 0).
mul_nonneg- The product of two non-negative numbers is non-negative (0 ≤ a → 0 ≤ b → 0 ≤ a * b).
- The non-linear arithmetic steps (lines 9 and 16) delegate polynomial identity checks and inequalities to the `nlinarith` decision procedure.
Written by a model (gemini-3.8-flash) from the Lean proof and the goal states Lean reported at each step. Lean verified the proof; this explanation is a reading aid and is not itself machine-checked.
Lean source view module on GitHub
Choose a tactic — a marked line — to see the goals Lean had before and after it, as in Lean’s infoview. Step with ← → or j k.
theorem parallel_mw_balancing_numerator_bound (a c A B C : ℝ) (ha : 1 ≤ a) (hc : 1 ≤ c) (hA : a + 1 ≤ A) (hC : c + 1 ≤ C) (hquad : a^2 + a ≤ 2*A) (hgap : a = 1 ∨ 2 ≤ a) (hsmall : a = 1 → A = 2) (hB : 2*A ≤ B) : B*(a*(a+2)+2*(a+1)*c) ≤ 2*(2*A*B*C-2*A-4*B-2*C+6) := by have hp := parallel_mw_balancing_base_polynomial a c A C ha hc hA hC hquad hgap hsmall have hApos : 0 < A := by linarith have hAC : 0 ≤ A+C-3 := by linarith have hAk : 0 ≤ A*(4*A*C-8-(a*(a+2)+2*(a+1)*c)) := by nlinarith have hk : 0 ≤ 4*A*C-8-(a*(a+2)+2*(a+1)*c) := by by_contra hh have hn : 4*A*C-8-(a*(a+2)+2*(a+1)*c) < 0 := by linarith have hh' := mul_neg_of_pos_of_neg hApos hn linarith have hm := mul_nonneg (show 0 ≤ B-2*A by linarith) hk nlinarith