Probability normalization for independent unit exponentials #
This module supplies the product-density normalization needed to turn the normalized-exponential integral identity into a probability-law statement.
theorem
Feige.integrable_pi_unitExponentialDensity
{n : ℕ}
:
MeasureTheory.Integrable (fun (x : Fin n → ℝ) => ∏ i : Fin n, unitExponentialDensity (x i)) MeasureTheory.volume
The product density of n independent unit exponentials is
integrable on the coordinate function space.
Real-valued joint density on ℝ × (Fin n → ℝ).
Equations
- Feige.jointUnitExponentialDensity e = Feige.unitExponentialDensity e.1 * ∏ i : Fin n, Feige.unitExponentialDensity (e.2 i)
Instances For
ENNReal form of normalization of the full independent exponential
product density.
theorem
Feige.lintegral_positiveExponentialOrthant_exp_neg_total
{n : ℕ}
:
∫⁻ (e : ℝ × (Fin n → ℝ)) in positiveExponentialOrthant, ENNReal.ofReal (Real.exp (-exponentialTotal e)) = 1
Exact Lebesgue volume of the standard full simplex.