Zulip Chat Archive
Stream: Is there code for X?
Topic: Filter.Tendsto.exists_isBounded_image
Jireh Loreaux (Nov 07 2023 at 19:52):
I can't find this, but perhaps my phrasing is just off. I can't imagine that we really don't have it though. Note: I'm aware of the existence of docs#Filter.Tendsto.isBoundedUnder_le, and while you could use that to get this, it's not really any easier.
import Mathlib.Topology.MetricSpace.Bounded
open Bornology Metric Topology
theorem Filter.Tendsto.exists_isBounded_image {α β : Type*} [PseudoMetricSpace β]
{l : Filter α} {f : α → β} {x : β} (hf : Tendsto f l (𝓝 x)) :
∃ s ∈ l, IsBounded (f '' s) := by
refine ⟨_, hf <| nhds_basis_ball.mem_of_mem zero_lt_one, ?_⟩
apply Metric.isBounded_ball (x := x) (r := 1) |>.subset
rintro - ⟨y, hy, rfl⟩
exact hy
Yury G. Kudryashov (Nov 08 2023 at 02:54):
This should follow from disjoint (nhds _) (cobounded _)
.
Yury G. Kudryashov (Nov 08 2023 at 02:57):
And docs#Filter.HasBasis.disjoint_iff_left
Yury G. Kudryashov (Nov 08 2023 at 02:58):
Something like (l.basis_sets.map _).disjoint_iff_left.1 ((disjoint_nhds_cobounded _).mono_left hf)
(not tested)
Yury G. Kudryashov (Nov 08 2023 at 02:58):
Probably with compl_mem_cobounded
Yury G. Kudryashov (Nov 08 2023 at 03:00):
The only reason I like this strangely looking proof more than yours is that we may want to add typeclasses for "topology plays nicely with bornology" later if we have some use cases.
Yury G. Kudryashov (Nov 08 2023 at 03:01):
And if we do this, it may be useful to see which basis properties (like "nhds are bounded") were used in the proof.
Last updated: Dec 20 2023 at 11:08 UTC