nae_sine_contraction_strong
For real u with 0 ≤ u ≤ π/12, (√8/3) sin²(6u) ≤ sin(5u).
Statement
theorem nae_sine_contraction_strong (u : ℝ) (hu0 : 0 ≤ u) (hu : u ≤ Real.pi / 12) :
Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2 ≤ Real.sin (5 * u)- Source
- arXiv:1211.0721, Section 3.3, p-computation rule analysis
- Verified
- 14 Nov 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 5
- nae_angle_polyThe real-valued function nae_angle_poly is defined for real x by 9(16x²−20x+5)²−128x(1−x)²(3−4x)⁴(1−4x)⁴.
- nae_angle_poly_nonnegFor real x with 0 ≤ x ≤ 1/14, nae_angle_poly x is nonnegative.
- nae_sin_fiveFor real u, sin(5u) = sin(u)(16 sin⁴(u) − 20 sin²(u) + 5).
- nae_sin_sixFor real u, sin(6u) = 2 sin(u) cos(u)(3−4 sin²(u))(1−4 sin²(u)).
- nae_sin_small_squareFor real u with 0 ≤ u ≤ π/12, sin²(u) ≤ 1/14.
- Used by 1
- nae_angle_contraction_strongFor real θ with 0 ≤ θ ≤ π, arccos(1−(4/9)(1−cos θ)²) ≤ 5θ/6.
The proof, explained
To prove that (√8/3) sin²(6u) ≤ sin(5u) for u ∈ [0, π/12], both sides are verified to be non-negative, reducing the claim to the comparison of their squares: (8/9) sin⁴(6u) ≤ sin²(5u). Expanding sin(5u) and sin(6u) with multiple-angle formulas and substituting x = sin²(u) shows that 9 sin²(5u) - 8 sin⁴(6u) factors as x * nae_angle_poly(x). Since 0 ≤ x ≤ 1/14 for u in this range, the auxiliary polynomial nae_angle_poly(x) is non-negative, which proves the squared inequality and completes the argument.
Setting x = sin²(u), we have 0 ≤ x ≤ 1/14, which implies that nae_angle_poly(x) ≥ 0
We introduce the substitution x = sin²(u). Being a square of a real number, x ≥ 0. Given 0 ≤ u ≤ π/12, the lemma nae_sin_small_square gives the upper bound x ≤ 1/14. Applying nae_angle_poly_nonneg to this interval establishes that nae_angle_poly(x) is non-negative.
uses
nae_sin_small_square,nae_angle_poly_nonneglet x : ℝ := Real.sin u ^ 2 have hx0 : 0 ≤ x := by dsimp [x]; positivity have hx : x ≤ 1 / 14 := nae_sin_small_square u hu0 hu have hpoly := nae_angle_poly_nonneg x hx0 hx
cos²(u) = 1 - sin²(u) and cos⁴(u) = (1 - sin²(u))²
From the Pythagorean identity sin²(u) + cos²(u) = 1 (Real.sin_sq_add_cos_sq), we express cos²(u) as 1 - sin²(u). Squaring both sides yields cos⁴(u) = (1 - sin²(u))².
uses
Real.sin_sq_add_cos_sqhave hcos2 : Real.cos u ^ 2 = 1 - Real.sin u ^ 2 := by nlinarith [Real.sin_sq_add_cos_sq u] have hcos4 : Real.cos u ^ 4 = (1 - Real.sin u ^ 2) ^ 2 := by calc Real.cos u ^ 4 = (Real.cos u ^ 2) ^ 2 := by ring _ = _ := by rw [hcos2]
The difference 9 sin²(5u) - 8 sin⁴(6u) equals x * nae_angle_poly(x), and is therefore non-negative
We expand sin(5u) and sin(6u) using their multiple-angle formulas nae_sin_five and nae_sin_six. Substituting cos⁴(u) = (1 - sin²(u))² and simplifying shows that 9 sin²(5u) - 8 sin⁴(6u) algebraically equals x * nae_angle_poly(x). Since x ≥ 0 and nae_angle_poly(x) ≥ 0, their product is non-negative by mul_nonneg.
uses
nae_sin_five,nae_sin_six,mul_nonneghave hfactor : 9 * Real.sin (5 * u) ^ 2 - 8 * Real.sin (6 * u) ^ 4 = x * nae_angle_poly x := by rw [nae_sin_five, nae_sin_six] dsimp [x, nae_angle_poly] calc _ = 9 * Real.sin u ^ 2 * (16 * Real.sin u ^ 4 - 20 * Real.sin u ^ 2 + 5) ^ 2 - 128 * Real.sin u ^ 4 * Real.cos u ^ 4 * (3 - 4 * Real.sin u ^ 2) ^ 4 * (1 - 4 * Real.sin u ^ 2) ^ 4 := by ring _ = _ := by rw [hcos4]; ring have hnonneg : 0 ≤ x * nae_angle_poly x := mul_nonneg hx0 hpoly
The inequality holds for the squared expressions: ((√8/3) sin²(6u))² ≤ sin²(5u)
Squaring the left-hand side gives ((√8/3) sin²(6u))² = (8/9) sin⁴(6u), using Real.sq_sqrt along with mul_pow and div_pow. Since 9 sin²(5u) - 8 sin⁴(6u) ≥ 0, dividing by 9 immediately yields (8/9) sin⁴(6u) ≤ sin²(5u).
uses
Real.sq_sqrt,mul_pow,div_powhave hsquared : (Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2) ^ 2 ≤ Real.sin (5 * u) ^ 2 := by have hsqrt : (Real.sqrt 8) ^ 2 = 8 := Real.sq_sqrt (by norm_num) have hid : (Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2) ^ 2 = (8 / 9 : ℝ) * Real.sin (6 * u) ^ 4 := by rw [mul_pow, div_pow, hsqrt]; norm_num; ring rw [hid] nlinarith
Both (√8/3) sin²(6u) and sin(5u) are non-negative
The left-hand side is non-negative as a product of non-negative factors. For the right-hand side, 0 ≤ u ≤ π/12 implies 0 ≤ 5u ≤ 5π/12 ≤ π (using Real.pi_pos). By Real.sin_nonneg_of_nonneg_of_le_pi, the sine of any angle in [0, π] is non-negative, so sin(5u) ≥ 0.
uses
Real.sin_nonneg_of_nonneg_of_le_pi,Real.pi_poshave hleft : 0 ≤ Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2 := by positivity have hright : 0 ≤ Real.sin (5 * u) := Real.sin_nonneg_of_nonneg_of_le_pi (by positivity) (by nlinarith [Real.pi_pos])
(√8/3) sin²(6u) ≤ sin(5u)
Because both sides are non-negative, the inequality of their squares implies the inequality itself by sq_le_sq₀.
uses
sq_le_sq₀exact (sq_le_sq₀ hleft hright).mp hsquared
nae_sin_small_square- For u with 0 ≤ u ≤ π/12, sin²(u) ≤ 1/14
nae_angle_poly_nonneg- For real x with 0 ≤ x ≤ 1/14, the polynomial nae_angle_poly(x) is non-negative
Real.sin_sq_add_cos_sq- For any real x, sin²(x) + cos²(x) = 1
nae_sin_five- Quintuple-angle identity: sin(5u) = sin(u) * (16 sin⁴(u) - 20 sin²(u) + 5)
nae_sin_six- Sextuple-angle identity: sin(6u) = 2 sin(u) cos(u) * (3 - 4 sin²(u)) * (1 - 4 sin²(u))
mul_nonneg- The product of two non-negative numbers is non-negative
Real.sq_sqrt- For x ≥ 0, (√x)² = x
mul_pow- Power distributes over multiplication: (a * b)ⁿ = aⁿ * bⁿ
div_pow- Power distributes over division: (a / b)ⁿ = aⁿ / bⁿ
Real.sin_nonneg_of_nonneg_of_le_pi- If 0 ≤ x ≤ π, then sin(x) ≥ 0
Real.pi_pos- π is strictly positive
sq_le_sq₀- For non-negative numbers a and b, a² ≤ b² is equivalent to a ≤ b
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 nae_sine_contraction_strong (u : ℝ) (hu0 : 0 ≤ u) (hu : u ≤ Real.pi / 12) : Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2 ≤ Real.sin (5 * u) := by let x : ℝ := Real.sin u ^ 2 have hx0 : 0 ≤ x := by dsimp [x]; positivity have hx : x ≤ 1 / 14 := nae_sin_small_square u hu0 hu have hpoly := nae_angle_poly_nonneg x hx0 hx have hcos2 : Real.cos u ^ 2 = 1 - Real.sin u ^ 2 := by nlinarith [Real.sin_sq_add_cos_sq u] have hcos4 : Real.cos u ^ 4 = (1 - Real.sin u ^ 2) ^ 2 := by calc Real.cos u ^ 4 = (Real.cos u ^ 2) ^ 2 := by ring _ = _ := by rw [hcos2] have hfactor : 9 * Real.sin (5 * u) ^ 2 - 8 * Real.sin (6 * u) ^ 4 = x * nae_angle_poly x := by rw [nae_sin_five, nae_sin_six] dsimp [x, nae_angle_poly] calc _ = 9 * Real.sin u ^ 2 * (16 * Real.sin u ^ 4 - 20 * Real.sin u ^ 2 + 5) ^ 2 - 128 * Real.sin u ^ 4 * Real.cos u ^ 4 * (3 - 4 * Real.sin u ^ 2) ^ 4 * (1 - 4 * Real.sin u ^ 2) ^ 4 := by ring _ = _ := by rw [hcos4]; ring have hnonneg : 0 ≤ x * nae_angle_poly x := mul_nonneg hx0 hpoly have hsquared : (Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2) ^ 2 ≤ Real.sin (5 * u) ^ 2 := by have hsqrt : (Real.sqrt 8) ^ 2 = 8 := Real.sq_sqrt (by norm_num) have hid : (Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2) ^ 2 = (8 / 9 : ℝ) * Real.sin (6 * u) ^ 4 := by rw [mul_pow, div_pow, hsqrt]; norm_num; ring rw [hid] nlinarith have hleft : 0 ≤ Real.sqrt 8 / 3 * Real.sin (6 * u) ^ 2 := by positivity have hright : 0 ≤ Real.sin (5 * u) := Real.sin_nonneg_of_nonneg_of_le_pi (by positivity) (by nlinarith [Real.pi_pos]) exact (sq_le_sq₀ hleft hright).mp hsquared