ose_cube_scalar_certificate
For real p at least two and u between zero and one, four times p cubed times u squared times one minus u is at most (p minus one) times (p squared minus u squared) squared.
Statement
theorem ose_cube_scalar_certificate (p u : ℝ) (hp : 2 ≤ p) (hu0 : 0 ≤ u) (hu1 : u ≤ 1) :
4 * p^3 * u^2 * (1-u) ≤ (p-1) * (p^2-u^2)^2- Source
- arXiv:2611.08546, Section 6.1; scalar certificate for the spherical embedding
- Verified
- 15 Nov 2026
- Axioms
Classical.choiceQuot.soundpropext- Used by 1
- ose_spherical_shrink_scalar_boundsUnder the small-s parameter assumptions for the spherical two-layer construction, the two radial scaling factors have the ordering and…
The proof, explained
The inequality is established by decomposing it into two independent bounds connected through an intermediate expression. First, the u-dependent term is bounded by 27·u²·(1 - u) ≤ 4 using the factorisation (3u - 2)²·(3u + 1) ≥ 0. Second, the p-dependent term is bounded by 16·p³ ≤ 27·(p - 1)·(p² - 1)², which is proved by expanding in the shifted variable t = p - 2 ≥ 0 and noting that all polynomial coefficients are positive. Finally, since u² ≤ 1 implies (p² - 1)² ≤ (p² - u²)², combining these certificates proves the result.
0 ≤ p and u² ≤ 1
Since p ≥ 2, we immediately have p ≥ 0. Since 0 ≤ u ≤ 1, squaring gives u² ≤ 1.
have hp0 : 0 ≤ p := by linarith have hu2 : u^2 ≤ 1 := by nlinarith
0 ≤ 4 - 27·u²·(1 - u)
We express 4 - 27·u²·(1 - u) as the product (3u - 2)²·(3u + 1). The square (3u - 2)² is non-negative by sq_nonneg, and 3u + 1 ≥ 0 since u ≥ 0. Hence their product is non-negative by mul_nonneg, which shows that 27·u²·(1 - u) ≤ 4.
uses
mul_nonneg,sq_nonneghave ha : 0 ≤ 4 - 27*u^2*(1-u) := by have hv : 0 ≤ (3*u-2)^2 * (3*u+1) := mul_nonneg (sq_nonneg _) (by linarith) nlinarith [hv]
For t = p - 2 ≥ 0, the polynomial 27·t⁵ + 243·t⁴ + 794·t³ + 1146·t² + 699·t + 115 is non-negative
We introduce the shifted variable t = p - 2, which satisfies t ≥ 0 since p ≥ 2. The quintic polynomial in t has strictly positive coefficients, so every term is non-negative for t ≥ 0.
let t := p - 2 have ht : 0 ≤ t := by dsimp [t]; linarith have hpol : 0 ≤ 27*t^5 + 243*t^4 + 794*t^3 + 1146*t^2 + 699*t + 115 := by positivity
0 ≤ 27·(p - 1)·(p² - 1)² - 16·p³
Substituting p = t + 2 and expanding 27·(p - 1)·(p² - 1)² - 16·p³ as a polynomial in t yields precisely the quintic polynomial with positive coefficients from the previous step. Since that polynomial is non-negative, the inequality holds.
have hb : 0 ≤ 27*(p-1)*(p^2-1)^2 - 16*p^3 := by have heq : 27*(p-1)*(p^2-1)^2 - 16*p^3 = 27*t^5 + 243*t^4 + 794*t^3 + 1146*t^2 + 699*t + 115 := by dsimp [t]; ring linarith
(p² - 1)² ≤ (p² - u²)²
Since p ≥ 2, p² - 1 ≥ 3 ≥ 0. Since u² ≤ 1, we have p² - 1 ≤ p² - u². Since both sides are non-negative, taking the second power preserves the inequality by pow_le_pow_left₀.
uses
pow_le_pow_left₀have hsq : (p^2-1)^2 ≤ (p^2-u^2)^2 := by apply pow_le_pow_left₀ (by nlinarith) (by nlinarith) 2
0 ≤ (p - 1)·((p² - u²)² - (p² - 1)²)
Since p ≥ 2, p - 1 ≥ 0. From the previous step, (p² - u²)² - (p² - 1)² ≥ 0 by sub_nonneg.mpr. The product of two non-negative terms is non-negative by mul_nonneg.
uses
mul_nonneg,sub_nonneg.mprhave hc : 0 ≤ (p-1) * ((p^2-u^2)^2 - (p^2-1)^2) := mul_nonneg (by linarith) (sub_nonneg.mpr hsq)
0 ≤ p³·(4 - 27·u²·(1 - u))
Since p ≥ 0, p³ is non-negative by pow_nonneg. Multiplying with the previously established bound 4 - 27·u²·(1 - u) ≥ 0 via mul_nonneg gives the claim.
uses
mul_nonneg,pow_nonneghave hd : 0 ≤ p^3 * (4 - 27*u^2*(1-u)) := mul_nonneg (pow_nonneg hp0 _) ha4·p³·u²·(1 - u) ≤ (p - 1)·(p² - u²)² automation
The desired inequality is verified by combining the three certificates: 27 times the u-replacement bound (hc), plus the univariate bound in p (hb), plus 4 times the u-maximization bound (hd). Summing these three non-negative terms cancels the intermediate expressions and leaves exactly 27·((p - 1)·(p² - u²)² - 4·p³·u²·(1 - u)) ≥ 0.
nlinarith [hb, hc, hd]
mul_nonneg- The product of two non-negative numbers is non-negative.
pow_nonneg- Any natural power of a non-negative number is non-negative.
pow_le_pow_left₀- For non-negative numbers a ≤ b, taking n-th powers preserves the inequality: a^n ≤ b^n.
sq_nonneg- The square of any real number is non-negative.
sub_nonneg.mpr- If b ≤ a, then a - b is non-negative.
- At the final step, nlinarith finds a positive linear combination of the three certificates (hb, hc, hd) that cancels the intermediate terms and leaves 27 times the desired inequality.
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 ose_cube_scalar_certificate (p u : ℝ) (hp : 2 ≤ p) (hu0 : 0 ≤ u) (hu1 : u ≤ 1) : 4 * p^3 * u^2 * (1-u) ≤ (p-1) * (p^2-u^2)^2 := by have hp0 : 0 ≤ p := by linarith have hu2 : u^2 ≤ 1 := by nlinarith have ha : 0 ≤ 4 - 27*u^2*(1-u) := by have hv : 0 ≤ (3*u-2)^2 * (3*u+1) := mul_nonneg (sq_nonneg _) (by linarith) nlinarith [hv] let t := p - 2 have ht : 0 ≤ t := by dsimp [t]; linarith have hpol : 0 ≤ 27*t^5 + 243*t^4 + 794*t^3 + 1146*t^2 + 699*t + 115 := by positivity have hb : 0 ≤ 27*(p-1)*(p^2-1)^2 - 16*p^3 := by have heq : 27*(p-1)*(p^2-1)^2 - 16*p^3 = 27*t^5 + 243*t^4 + 794*t^3 + 1146*t^2 + 699*t + 115 := by dsimp [t]; ring linarith have hsq : (p^2-1)^2 ≤ (p^2-u^2)^2 := by apply pow_le_pow_left₀ (by nlinarith) (by nlinarith) 2 have hc : 0 ≤ (p-1) * ((p^2-u^2)^2 - (p^2-1)^2) := mul_nonneg (by linarith) (sub_nonneg.mpr hsq) have hd : 0 ≤ p^3 * (4 - 27*u^2*(1-u)) := mul_nonneg (pow_nonneg hp0 _) ha nlinarith [hb, hc, hd]