Zulip Chat Archive
Stream: mathlib4
Topic: Finding and importing notation ℕ
Bob Rubbens (Sep 23 2025 at 06:50):
Hi everyone,
I'm trying to use the notation ℕ for natural numbers. I found through a random topic about mathlib that the following import does the trick:
import Mathlib.Data.Real.Basic
I've tried searching for "ℕ" on the auto generated mathlib docs, but that only yielded some unrelated definitions (I think). Any time I click on the type "ℕ" in the mathlib documentation, I am redirected to the definition of Nat in the standard library.
My questions are:
- What is the precise import I need to get the "ℕ" notation to work? I think the import I have now also imports a bunch of Real stuff, which is fine, but I'm curious about where the exact definition of this notation is.
- How could I conceivably have found this notation on my own. Are there any other sources or search methods I should consider when I have questions like this?
Kevin Buzzard (Sep 23 2025 at 06:53):
If I right-click on ℕ and then choose "go to declaration" I see that the notation is defined in Mathlib.Data.Nat.Notation.
Damiano Testa (Sep 23 2025 at 07:54):
You can also do this:
import Mathlib
/--
info: Found 4 uses among 2394 syntax declarations
In `Mathlib.Data.ENat.Defs`:
«termℕ∞»: 'ℕ∞'
In `Mathlib.Data.Nat.Notation`:
termℕ: 'ℕ'
In `Mathlib.Data.PNat.Notation`:
«termℕ+»: 'ℕ+'
In `Mathlib.Topology.Category.LightProfinite.Sequence`:
LightProfinite.«termℕ∪{∞}»: 'ℕ∪{∞}'
-/
#guard_msgs in
#find_syntax "ℕ"
Violeta Hernández (Sep 24 2025 at 07:58):
we have notation ℕ∪{∞}? yikes
Kenny Lau (Sep 24 2025 at 11:50):
i'm in the camp of "as much notation as possible"
Alex Meiburg (Sep 24 2025 at 18:20):
I am too, but that is not what I would have expected ℕ∪{∞} to mean..
Aaron Liu (Sep 24 2025 at 19:58):
it's ENat, as a bundled second countable totally disconnected compact hausdorff topological space
Dagur Asgeirsson (Sep 24 2025 at 20:15):
It's specialized notation for condensed mathematics, living inside the LightProfinite namespace :shrug:
Floris van Doorn (Sep 25 2025 at 18:39):
Kenny Lau said:
i'm in the camp of "as much notation as possible"
Agreed, on the condition that it is scoped.
Kenny Lau (Sep 25 2025 at 19:13):
Floris van Doorn said:
Agreed, on the condition that it is scoped.
Agreed, on the condition that docgen opens all scoped notations, seeing TensorProduct and CategoryStruct.comp on the docs doesn't really help anyone
Last updated: Dec 20 2025 at 21:32 UTC