Zulip Chat Archive

Stream: maths

Topic: Holomorph of a group


Seewoo Lee (Oct 28 2025 at 18:43):

For a group GG, the holomorph Hol(G)\mathrm{Hol}(G) is simply defined as the semidirect product GAut(G)G \rtimes \mathrm{Aut}(G). I recently saw the definition from Dummit--Foote and realized that it does not exist in mathlib yet. It is fairly easy to formalize the definition (since we already have semidirect product) but not sure if it worths to add assuming I don't have any nice theorems / applications using it. (Dummit--Foote include some computations for small GG.) What's usual practice in such case?

Kenny Lau (Oct 28 2025 at 19:20):

@Seewoo Lee I don't speak for mathlib, but I think usually mathlib wouldn't include it in this case.

Ruben Van de Velde (Oct 28 2025 at 22:43):

I guess it depends - I don't think you need Big Results:trademark:, but you'd be expected to have at least some basic lemmas about the definition so you can actually use it

Seewoo Lee (Oct 28 2025 at 22:48):

I see, probably I will think of some basic facts (things on wikipedia or textbooks). There are some big(?) theorems using it but haven’t looked in details.

Yury G. Kudryashov (Oct 29 2025 at 01:23):

At least, we can introduce it as an abbrev so that loogle can find it.

Kenny Lau (Oct 29 2025 at 08:08):

is that how loogle works?

Scott Carnahan (Oct 30 2025 at 13:45):

A basic fact you could try formalizing is that, as a subgroup of Perm(G), the holomorph is equal to the normalizer of the left regular subgroup.

Kenny Lau (Oct 30 2025 at 13:51):

how do you embed it into Perm(G)?

Scott Carnahan (Oct 30 2025 at 19:16):

G acts by left multiplication, and Aut(G) acts by automorphisms.

Kenny Lau (Oct 30 2025 at 19:21):

I think this map already should be in mathlib (cc @Seewoo Lee)


Last updated: Dec 20 2025 at 21:32 UTC