Documentation

LeanPool.ComputableReal.SpecialFunctions.Pi

Verified rational bounds for pi #

Rational lower and upper bounds for Real.pi are derived from the Real.sqrtTwoAddSeries approximation, shown to converge, and packaged into a ComputableℝSeq.Pi, giving an IsComputable instance for Real.pi. The bounds go through the noncomputable square-root sequences, so Pi and the derived bounds piLb/piUb are noncomputable Lean terms.

@[instance_reducible]
Equations
  • One or more equations did not get rendered due to their size.
theorem ComputableℝSeq.sqrtTwoAddSeriesN_bounds (n k : ℕ) (hk : 3 ≤ k) :
↑(sqrtTwoAddSeriesN n).ub k ≤ ↑(sqrtTwoAddSeriesN n).lb k + 18 * ↑n / 2 ^ k
theorem ComputableℝSeq.sqrtTwoSubSqrtTwoAddSeries_lb (n k : ℕ) (hk : 3 ≤ k) :
√(2 - Real.sqrtTwoAddSeries 0 n) ≤ ↑(↑(sqrtTwoSubSqrtTwoAddSeriesN n).lb k) + (18 * ↑n * 2 ^ n + 4) / 2 ^ k
theorem ComputableℝSeq.sqrtTwoSubSqrtTwoAddSeries_ub (n k : ℕ) (hk : 3 ≤ k) :
↑(↑(sqrtTwoSubSqrtTwoAddSeriesN n).ub k) - (18 * ↑n * 2 ^ n + 14) / 2 ^ k ≤ √(2 - Real.sqrtTwoAddSeries 0 n)
noncomputable def ComputableℝSeq.piLb (n : ℕ) :

See theorem Real.pi_lt_sqrtTwoAddSeries in Mathlib

Equations
Instances For
    noncomputable def ComputableℝSeq.piUb (n : ℕ) :

    See theorem Real.pi_gt_sqrtTwoAddSeries in Mathlib

    Equations
    Instances For
      theorem ComputableℝSeq.piLb_ge_pi_sub_pow (n : ℕ) (hn : 0 < n) :
      Real.pi - 41 * ↑n / 2 ^ n ≤ ↑(piLb n)
      theorem ComputableℝSeq.piUb_le_pi_add_pow (n : ℕ) (hn : 0 < n) :
      ↑(piUb n) ≤ Real.pi + 51 * ↑n / 2 ^ n

      Definition of Pi.

      Equations
      • One or more equations did not get rendered due to their size.
      Instances For