Chebyshev polynomials over the reals: orthogonality #
Chebyshev T polynomials are orthogonal with respect to √(1 - x ^ 2)⁻¹.
Main statements #
- integrable_measureT: continuous functions are integrable with respect to Lebesgue measure
scaled by
√(1 - x ^ 2)⁻¹and restricted to(-1, 1]. - integral_eval_T_real_mul_evalT_real_measureT_of_ne:
if
n ≠ mthen the integral ofT_n * T_mequals0. - integral_eval_T_real_mul_self_measureT_zero:
if
n = m = 0then the integral equalsπ. - integral_eval_T_real_mul_self_measureT_of_ne_zero:
if
n = m ≠ 0then the integral equalsπ / 2.
TODO #
- Prove that Chebyshev U polynomials are orthogonal with respect to
√(1 - x ^ 2) - Bundle Chebyshev T polynomials into a HilbertBasis for MeasureTheory.Lp ℝ 2 measureT
Lebesgue measure scaled by √(1 - x ^ 2)⁻¹.
Equations
- Polynomial.Chebyshev.measureT = (MeasureTheory.volume.withDensity fun (x : ℝ) => ↑⟨√(1 - x ^ 2)⁻¹, ⋯⟩).restrict (Set.Ioc (-1) 1)
Instances For
theorem
Polynomial.Chebyshev.intervalIntegrable_sqrt_one_sub_sq_inv :
IntervalIntegrable (fun (x : ℝ) => √(1 - x ^ 2)⁻¹) MeasureTheory.volume (-1) 1
theorem
Polynomial.Chebyshev.integrable_measureT
{f : ℝ → ℝ}
(hf : ContinuousOn f (Set.Icc (-1) 1))
:
@[deprecated Polynomial.Chebyshev.integral_measureT_eq_integral_cos (since := "2026-03-19")]
Alias of Polynomial.Chebyshev.integral_measureT_eq_integral_cos.