Zulip Chat Archive

Stream: Is there code for X?

Topic: Semialgebraic Sets


Tomaz Mascarenhas (Apr 22 2025 at 17:35):

Do we have anything on Semialgebraic sets?

Kevin Buzzard (Apr 22 2025 at 17:55):

Not even the definition, as far as I know.

Kevin Buzzard (Apr 22 2025 at 17:55):

Proving o-minimality would be a really cool project!

Yaël Dillies (Apr 22 2025 at 17:56):

It also is tangentially related to my research :smile:

Tomaz Mascarenhas (Apr 22 2025 at 19:41):

I see; is there any structure in mathlib that is constructed in a similar way? i.e. starting with a property (P(x) > 0 in this case) and then taking arbitrary conjunctions of these (forming basic semialgebraic sets) and then taking arbitrary disjunctions of these?

Aaron Liu (Apr 22 2025 at 19:44):

It seems to be only finite unions and intersections?

Tomaz Mascarenhas (Apr 22 2025 at 19:44):

Yes I think so

Aaron Liu (Apr 22 2025 at 19:45):

otherwise you could do anything as a union of singletons

Aaron Liu (Apr 22 2025 at 19:48):

Well docs#TopologicalSpace.generateFrom seems pretty similar

Aaron Liu (Apr 22 2025 at 19:50):

As well as the ones for filter and measure space and stuff

Aaron Liu (Apr 22 2025 at 19:50):

We also have docs#FreeRing, which is sums of products of elements

Fabrizio Barroero (Apr 22 2025 at 22:41):

I think @Johan Commelin once told me he had some project on this in lean 3. I can’t remember if it was about semi algebraic sets or o-minimality.

Adam Topaz (Apr 22 2025 at 22:56):

IMO the proper way to work with semialgebraic sets is to consider them as definable sets in the appropriate language w.r.t. appropriate kinds of first-order formulas.

Adam Topaz (Apr 22 2025 at 22:59):

(The theory of real closed fields has quantifier elimination)

Tomaz Mascarenhas (Apr 22 2025 at 23:43):

Adam Topaz said:

IMO the proper way to work with semialgebraic sets is to consider them as definable sets in the appropriate language w.r.t. appropriate kinds of first-order formulas.

You mean as defined here?

https://github.com/leanprover-community/mathlib4/blob/04c82c164d0b74dbfe3fcd0d7e174938d1112399/Mathlib/ModelTheory/Definability.lean

Luigi Massacci (Apr 23 2025 at 14:26):

You can look here for a purely geometric approach: https://github.com/rwbarton/lean-omin by @Reid Barton and @Johan Commelin

Johan Commelin (Apr 24 2025 at 18:33):

Just seeing this thread now. Thanks for posting that link Luigi!

@Niels Keukens @Lily <3 and @Joos Munneke are currently also working on a student project to formalize the basics of model-theoretic approach to o-minimality.


Last updated: May 02 2025 at 03:31 UTC