Zulip Chat Archive
Stream: mathlib4
Topic: Naming congruence of maps
Yaël Dillies (May 26 2025 at 10:26):
There is a general pattern that if A →something B is a type of maps and e : A ≃something A' is an isomorphism, then we get an isomorphism (A →something B) ≃something (A' →something B), and similarly for e : B ≃something B'
Yaël Dillies (May 26 2025 at 10:27):
Naming of these "congruence" isomorphisms is all over the place, and it would be good to unify them
Yaël Dillies (May 26 2025 at 10:28):
Eric and I are adding a few in #25142 and #25182, and we would like to know what the community thinks before we commit to more names
Yaël Dillies (May 26 2025 at 10:29):
Here are some examples (taken from Eric's PR):
- docs#AlternatingMap.domLCongr
- docs#ContinuousLinearEquiv.compContinuousMultilinearMapL
- docs#ContinuousLinearEquiv.continuousAlternatingMapCongrLeft
- docs#ContinuousLinearEquiv.continuousAlternatingMapCongrLeftEquiv (not a linear equiv)
- docs#ContinuousLinearEquiv.continuousMultilinearMapCongrRight
- docs#ContinuousLinearEquiv.continuousAlternatingMapCongrRight
Yaël Dillies (May 26 2025 at 10:30):
and some examples coming from my PR instead:
Eric Wieser (May 26 2025 at 10:31):
Some more examples:
- docs#LinearEquiv.congrLeft
- docs#LinearEquiv.funCongrLeft (the two
somethings are not from the same category here)
Eric Wieser (May 27 2025 at 11:31):
It looks like maybe FooEquiv.barMapCongrSide is the plurality winner right now, perhaps with some exception that says FooEquiv.fooMapCongrSide becomes FooEquiv.congrSide?
Eric Wieser (May 27 2025 at 11:34):
#24159 seems to support this (it renamed one of the items on your list)
Last updated: Dec 20 2025 at 21:32 UTC