cubic_poly_strict_mono_of_minors_lt_one
For any real numbers x, y, z satisfying x^2 < 1, y^2 < 1, and (x, y, z) ≠ (0, 0, 0), the cubic polynomial function mapping q to 1 + (x^2 + y^2) * q + 2 * x * y * z * q^2 + z^2 * q^3 is strictly increasing on ℝ.
Statement
theorem cubic_poly_strict_mono_of_minors_lt_one (x y z : ℝ) (hx : x ^ 2 < 1) (hy : y ^ 2 < 1) (hne : (x, y, z) ≠ (0, 0, 0)) : StrictMono (fun q : ℝ => 1 + (x ^ 2 + y ^ 2) * q + 2 * x * y * z * q ^ 2 + z ^ 2 * q ^ 3)
- Formalizes
- arXiv:2611.14694, Proof of Proposition 5.1 -- Real positive definite counterexamples to a monotonicity conjecture for the q-permanent
- Verified
- 16 Nov 2026
- Axioms
Classical.choiceQuot.soundpropext- Used by 1
- matrix_q_permanent_strict_mono_of_order_le_threeIf A is a non-diagonal real positive definite matrix of order n ≤ 3, then the function mapping q to the q-permanent of A is strictly…
Read back from the Lean
For all real numbers $x, y, z$, if $x^2 < 1$, $y^2 < 1$, and $(x, y, z) \neq (0, 0, 0)$, then the polynomial function $f : \mathbb{R} \to \mathbb{R}$ defined by $f(q) = 1 + (x^2 + y^2)q + 2xyz q^2 + z^2 q^3$ is strictly monotone (strictly increasing: for all $a, b \in \mathbb{R}$ with $a < b$, $f(a) < f(b)$).
Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.
The proof, explained
To prove that the cubic polynomial $P(q) = 1 + (x^2 + y^2)q + 2xyz q^2 + z^2 q^3$ is strictly increasing, it suffices to show that for any $q_1 < q_2$, the difference quotient $(P(q_2) - P(q_1))/(q_2 - q_1) = (x^2 + y^2) + 2xyz(q_1 + q_2) + z^2(q_1^2 + q_1 q_2 + q_2^2)$ is strictly positive. When $z = 0$, this expression simplifies to $x^2 + y^2$, which is strictly positive because $(x, y, z) \\neq (0, 0, 0)$. When $z \\neq 0$, completing the square expresses 12 times this quotient as the sum $(3z(q_1 + q_2) + 4xy)^2 + 4(3(x^2 + y^2) - 4x^2 y^2) + 3z^2(q_2 - q_1)^2$. The middle term is non-negative because $x^2 < 1$ and $y^2 < 1$, while the third term is strictly positive because $z \\neq 0$ and $q_1 \\neq q_2$, ensuring the quotient is strictly positive and thus $P(q_1) < P(q_2)$.
Let $q_1, q_2 \\in \\mathbb{R}$ satisfy $q_1 < q_2$.
To prove that the polynomial function $P(q) = 1 + (x^2 + y^2)q + 2xyz q^2 + z^2 q^3$ is strictly increasing, we introduce two arbitrary real numbers $q_1 < q_2$ and must show $P(q_1) < P(q_2)$.
intro q1 q2 hltWhen $z = 0$, the quadratic factor $(x^2 + y^2) + 2xyz(q_1 + q_2) + z^2(q_1^2 + q_1 q_2 + q_2^2)$ is strictly positive.
We begin proving that this quadratic factor is strictly positive, splitting into cases on whether $z = 0$. When $z = 0$, the expression simplifies to $x^2 + y^2$. If this sum of squares were non-positive, non-negativity of squares (`sq_nonneg`) would force $x = 0$ and $y = 0$, which together with $z = 0$ contradicts the hypothesis $(x, y, z) \\neq (0, 0, 0)$. Hence $x^2 + y^2 > 0$, establishing the claim for $z = 0$.
uses
sq_nonneghave hB : 0 < (x ^ 2 + y ^ 2) + 2 * x * y * z * (q1 + q2) + z ^ 2 * (q1 ^ 2 + q1 * q2 + q2 ^ 2) := by by_cases hz : z = 0 · have hApos : 0 < x ^ 2 + y ^ 2 := by by_contra h have hsum : x ^ 2 + y ^ 2 = 0 := by have hnonneg : 0 ≤ x ^ 2 + y ^ 2 := by positivity linarith have hx0 : x = 0 := by nlinarith [sq_nonneg x] have hy0 : y = 0 := by nlinarith [sq_nonneg y] apply hne simp [hx0, hy0, hz] simpa [hz] using hApos
$3(x^2 + y^2) - 4x^2 y^2 \\ge 0$.
Since $x^2 < 1$ and $y^2 < 1$, both $1 - y^2$ and $1 - x^2$ are non-negative. Using `mul_nonneg` with `sq_nonneg`, the products $x^2(1 - y^2)$ and $y^2(1 - x^2)$ are non-negative. Combining these inequalities via non-linear arithmetic shows that $3(x^2 + y^2) - 4x^2 y^2 \\ge 0$.
uses
mul_nonneg,sq_nonneg· have hkey : 0 ≤ 3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2 := by have h1 : 0 ≤ x ^ 2 * (1 - y ^ 2) := by exact mul_nonneg (sq_nonneg x) (by linarith) have h2 : 0 ≤ y ^ 2 * (1 - x ^ 2) := by exact mul_nonneg (sq_nonneg y) (by linarith) nlinarith
The quadratic factor $(x^2 + y^2) + 2xyz(q_1 + q_2) + z^2(q_1^2 + q_1 q_2 + q_2^2)$ is strictly positive in all cases.
When $z \\neq 0$, completing the square reveals that 12 times this factor equals $(3z(q_1 + q_2) + 4xy)^2 + 4(3(x^2 + y^2) - 4x^2 y^2) + 3z^2(q_2 - q_1)^2$. The first term is a square, hence non-negative by `sq_nonneg`. The second term is non-negative by the inequality established in the previous step. Because $z \\neq 0$ and $q_1 < q_2$, both $z^2$ and $(q_2 - q_1)^2$ are strictly positive by `sq_pos_of_ne_zero`, making the third term strictly positive. The entire sum is therefore strictly positive, which implies the factor itself is strictly positive.
uses
sq_nonneg,sq_pos_of_ne_zerohave h12 : 12 * ((x ^ 2 + y ^ 2) + 2 * x * y * z * (q1 + q2) + z ^ 2 * (q1 ^ 2 + q1 * q2 + q2 ^ 2)) = (3 * z * (q1 + q2) + 4 * x * y) ^ 2 + 4 * (3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2) + 3 * z ^ 2 * (q2 - q1) ^ 2 := by ring have hz2 : 0 < z ^ 2 := sq_pos_of_ne_zero hz have hd2 : 0 < (q2 - q1) ^ 2 := by apply sq_pos_of_ne_zero linarith have hlast : 0 < 3 * z ^ 2 * (q2 - q1) ^ 2 := by positivity have hsq : 0 ≤ (3 * z * (q1 + q2) + 4 * x * y) ^ 2 := sq_nonneg _ have hres : 0 ≤ 4 * (3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2) := by positivity have hrhs : 0 < (3 * z * (q1 + q2) + 4 * x * y) ^ 2 + 4 * (3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2) + 3 * z ^ 2 * (q2 - q1) ^ 2 := by linarith nlinarith
The difference $P(q_2) - P(q_1)$ factors as $(q_2 - q_1)((x^2 + y^2) + 2xyz(q_1 + q_2) + z^2(q_1^2 + q_1 q_2 + q_2^2))$.
Expanding both sides shows that subtracting the polynomial value at $q_1$ from that at $q_2$ factors into the difference $q_2 - q_1$ times the quadratic expression analyzed above.
have hdiff : (1 + (x ^ 2 + y ^ 2) * q2 + 2 * x * y * z * q2 ^ 2 + z ^ 2 * q2 ^ 3) - (1 + (x ^ 2 + y ^ 2) * q1 + 2 * x * y * z * q1 ^ 2 + z ^ 2 * q1 ^ 3) = (q2 - q1) * ((x ^ 2 + y ^ 2) + 2 * x * y * z * (q1 + q2) + z ^ 2 * (q1 ^ 2 + q1 * q2 + q2 ^ 2)) := by ring
$P(q_1) < P(q_2)$, completing the proof of strict monotonicity.
Because $q_1 < q_2$, the difference $q_2 - q_1$ is strictly positive. Multiplying this by the strictly positive quadratic factor via `mul_pos` shows that $P(q_2) - P(q_1) > 0$, from which linear arithmetic yields $P(q_1) < P(q_2)$.
uses
mul_poshave hpos : 0 < (1 + (x ^ 2 + y ^ 2) * q2 + 2 * x * y * z * q2 ^ 2 + z ^ 2 * q2 ^ 3) - (1 + (x ^ 2 + y ^ 2) * q1 + 2 * x * y * z * q1 ^ 2 + z ^ 2 * q1 ^ 3) := by rw [hdiff] exact mul_pos (by linarith) hB linarith
sq_nonneg- The square of any real number is non-negative ($0 \\le a^2$).
mul_nonneg- The product of two non-negative real numbers is non-negative.
sq_pos_of_ne_zero- The square of any non-zero real number is strictly positive ($a \\neq 0 \\implies 0 < a^2$).
mul_pos- The product of two strictly positive real numbers is strictly positive.
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.
/-- The cubic polynomial 1 + (x^2 + y^2)q + 2xyz q^2 + z^2 q^3 is strictly increasing on ℝ whenever x^2 < 1, y^2 < 1, and (x, y, z) ≠ (0, 0, 0). -/theorem cubic_poly_strict_mono_of_minors_lt_one (x y z : ℝ) (hx : x ^ 2 < 1) (hy : y ^ 2 < 1) (hne : (x, y, z) ≠ (0, 0, 0)) : StrictMono (fun q : ℝ => 1 + (x ^ 2 + y ^ 2) * q + 2 * x * y * z * q ^ 2 + z ^ 2 * q ^ 3) := by intro q1 q2 hlt have hB : 0 < (x ^ 2 + y ^ 2) + 2 * x * y * z * (q1 + q2) + z ^ 2 * (q1 ^ 2 + q1 * q2 + q2 ^ 2) := by by_cases hz : z = 0 · have hApos : 0 < x ^ 2 + y ^ 2 := by by_contra h have hsum : x ^ 2 + y ^ 2 = 0 := by have hnonneg : 0 ≤ x ^ 2 + y ^ 2 := by positivity linarith have hx0 : x = 0 := by nlinarith [sq_nonneg x] have hy0 : y = 0 := by nlinarith [sq_nonneg y] apply hne simp [hx0, hy0, hz] simpa [hz] using hApos · have hkey : 0 ≤ 3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2 := by have h1 : 0 ≤ x ^ 2 * (1 - y ^ 2) := by exact mul_nonneg (sq_nonneg x) (by linarith) have h2 : 0 ≤ y ^ 2 * (1 - x ^ 2) := by exact mul_nonneg (sq_nonneg y) (by linarith) nlinarith have h12 : 12 * ((x ^ 2 + y ^ 2) + 2 * x * y * z * (q1 + q2) + z ^ 2 * (q1 ^ 2 + q1 * q2 + q2 ^ 2)) = (3 * z * (q1 + q2) + 4 * x * y) ^ 2 + 4 * (3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2) + 3 * z ^ 2 * (q2 - q1) ^ 2 := by ring have hz2 : 0 < z ^ 2 := sq_pos_of_ne_zero hz have hd2 : 0 < (q2 - q1) ^ 2 := by apply sq_pos_of_ne_zero linarith have hlast : 0 < 3 * z ^ 2 * (q2 - q1) ^ 2 := by positivity have hsq : 0 ≤ (3 * z * (q1 + q2) + 4 * x * y) ^ 2 := sq_nonneg _ have hres : 0 ≤ 4 * (3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2) := by positivity have hrhs : 0 < (3 * z * (q1 + q2) + 4 * x * y) ^ 2 + 4 * (3 * (x ^ 2 + y ^ 2) - 4 * x ^ 2 * y ^ 2) + 3 * z ^ 2 * (q2 - q1) ^ 2 := by linarith nlinarith have hdiff : (1 + (x ^ 2 + y ^ 2) * q2 + 2 * x * y * z * q2 ^ 2 + z ^ 2 * q2 ^ 3) - (1 + (x ^ 2 + y ^ 2) * q1 + 2 * x * y * z * q1 ^ 2 + z ^ 2 * q1 ^ 3) = (q2 - q1) * ((x ^ 2 + y ^ 2) + 2 * x * y * z * (q1 + q2) + z ^ 2 * (q1 ^ 2 + q1 * q2 + q2 ^ 2)) := by ring have hpos : 0 < (1 + (x ^ 2 + y ^ 2) * q2 + 2 * x * y * z * q2 ^ 2 + z ^ 2 * q2 ^ 3) - (1 + (x ^ 2 + y ^ 2) * q1 + 2 * x * y * z * q1 ^ 2 + z ^ 2 * q1 ^ 3) := by rw [hdiff] exact mul_pos (by linarith) hB linarith