Properties of the integral of mulExpNegMulSq
#
The mapping mulExpNegMulSq
can be used to transform a function g : E → ℝ
into a bounded
function mulExpNegMulSq ε ∘ g : E → ℝ = fun x => g x * Real.exp (-ε * g x * g x)
. This file
contains results on the integral of mulExpNegMulSq g ε
with respect to a finite measure P
.
Lemmas #
tendsto_integral_mulExpNegMulSq_comp
: By the dominated convergence theorem andmulExpNegMulSq_abs_le_norm
, the integral ofmulExpNegMulSq ε ∘ g
with respect to a finite measureP
converges to the integral ofg
, asε → 0
;tendsto_integral_mul_one_plus_inv_smul_sq_pow
: The integral ofmulExpNegMulSq ε ∘ g
with respect to a finite measureP
can be approximated by the integral of the sequence approximating the exponential function,fun x => (g * (1 + (n : ℝ)⁻¹ • -(ε • g * g)) ^ n) x
. This allows to transfer properties of a subalgebra of functions containingg
to the functionmulExpNegMulSq ε ∘ g
, see e.g.integral_mulExpNegMulSq_comp_eq
.
Main Result #
dist_integral_mulExpNegMulSq_comp_le
: For a subalgebra of functions A
, if for any g ∈ A
the
integral with respect to two finite measures P, P'
coincide, then the difference of the integrals
of mulExpNegMulSq ε ∘ g
with respect to P, P'
is bounded by 6 * sqrt ε
.
This will be a key ingredient in the proof of theorem ext_of_forall_mem_subalgebra_integral_eq
(future work), where it is shown that a subalgebra of functions that separates points separates
finite measures.
The integral of mulExpNegMulSq ε ∘ g
with respect to a finite measure P
converges to the
integral of g
, as ε → 0
from above.
The integral of mulExpNegMulSq ε ∘ g
with respect to a finite measure P
can be
approximated by the integral of the sequence approximating the exponential function.
If for any g ∈ A
the integrals with respect to two finite measures P, P'
coincide, then the
difference of the integrals of mulExpNegMulSq ε ∘ g
with respect to P, P'
is bounded by
6 * sqrt ε
.