Stability of centered Gaussian differences #
The characteristic function proves that the difference of two independent copies, divided by sqrt(2), has the original law. No density assumption is needed, so singular covariance matrices are included.
Equations
- One or more equations did not get rendered due to their size.
Instances For
theorem
Verification.gaussianDifference_apply
(p : EuclideanSpace ℝ (Fin 2) × EuclideanSpace ℝ (Fin 2))
:
theorem
Verification.map_gaussianDifference
(r : ℝ)
:
have μ := ProbabilityTheory.multivariateGaussian 0 (bivariateCorrelation r);
MeasureTheory.Measure.map (⇑gaussianDifference) (μ.prod μ) = μ