theorem
proved
mollified_continuous
show as:
view math explainer →
open explainer
Read the cached plain-language explainer.
open lean source
IndisputableMonolith.Foundation.GeneralizedDAlembert on GitHub at line 218.
browse module
All declarations in this module, on Recognition.
explainer page
depends on
formal source
215
216/-- The normalized bump has compact support and is continuous; convolving
217with a continuous function leaves a continuous function. -/
218theorem mollified_continuous (G : ℝ → ℝ) (hG : Continuous G)
219 (φ : ContDiffBump (0 : ℝ)) : Continuous (mollified G φ) := by
220 unfold mollified
221 refine HasCompactSupport.continuous_convolution_left
222 (L := ContinuousLinearMap.lsmul ℝ ℝ) ?_ ?_ ?_
223 · exact φ.hasCompactSupport_normed (μ := (volume : MeasureTheory.Measure ℝ))
224 · have h0 : ContDiff ℝ ((0 : ℕ∞) : WithTop ℕ∞)
225 (φ.normed (μ := (volume : MeasureTheory.Measure ℝ))) :=
226 φ.contDiff_normed (μ := (volume : MeasureTheory.Measure ℝ)) (n := 0)
227 -- ContDiff at level 0 is continuity.
228 exact (contDiff_zero.mp (by exact_mod_cast h0))
229 · exact hG.locallyIntegrable (μ := (volume : MeasureTheory.Measure ℝ))
230
231/-- As the bump support shrinks, the mollification converges pointwise to
232the original continuous function. -/
233theorem mollified_pointwise_tendsto {ι : Type*}
234 (G : ℝ → ℝ) (hG : Continuous G)
235 {φ : ι → ContDiffBump (0 : ℝ)} {l : Filter ι}
236 (hφ : Filter.Tendsto (fun i => (φ i).rOut) l (nhds 0)) (x₀ : ℝ) :
237 Filter.Tendsto (fun i => mollified G (φ i) x₀) l (nhds (G x₀)) := by
238 unfold mollified
239 exact ContDiffBump.convolution_tendsto_right_of_continuous hφ hG x₀
240
241/-- Helper predicate naming the residual analytic content for Piece 1.
242
243`MollifierCkRoute` says: there is a parametrized family of `ContDiffBump`
244kernels with shrinking support. This is *infrastructure only* — the
245analytic blocker is uniform compact-set derivative bounds on
246`mollified G (φ n)`, not the existence of the bump family. -/
247def MollifierCkRoute : Prop :=
248 ∃ (φ : ℕ → ContDiffBump (0 : ℝ)),