Zulip Chat Archive
Stream: maths
Topic: p-norms on Matrices
Hououin Kyouma (Sep 06 2025 at 13:08):
image.png
Are there any pre-defined matrix norms in mathlib4 as defined in the figure? Or how should I define it?
Eric Wieser (Sep 06 2025 at 13:23):
We have this only for at the moment
Eric Wieser (Sep 06 2025 at 13:24):
The easiest path would be to generalize docs#Matrix.toEuclideanLin away from (which is trivial, you could call it Matrix.toLpLin), and then use the operator norm which is already in mathlib
Eric Wieser (Sep 06 2025 at 13:26):
docs#ContinuousLinearMap.sSup_sphere_eq_norm is then the theorem (skipping the middle part) for your equation
Hououin Kyouma (Sep 07 2025 at 10:49):
Eric Wieser said:
docs#ContinuousLinearMap.sSup_sphere_eq_norm is then the theorem (skipping the middle part) for your equation
[image.png](/user_uploads/3121/yu3gcTkHf5ph4LJyzqRieNaL/image.png)
Hououin Kyouma (Sep 07 2025 at 10:50):
I cannot find this theorem on mathlib, but I have found sSup_sphere_eq_norm. Should I use this theorem?
Eric Wieser (Sep 07 2025 at 11:13):
I don't know what you mean by "this" (the image in that message is misformatted).
Hououin Kyouma (Sep 07 2025 at 11:19):
theorem sSup_sphere_eq_norm [NormedAlgebra ℝ 𝕜] (f : E →SL[σ₁₂] F) :
sSup ((fun x => ‖f x‖) '' Metric.sphere 0 1) = ‖f‖ := by
simpa only [NNReal.coe_sSup, Set.image_image] using NNReal.coe_inj.2 f.sSup_sphere_eq_nnnorm
Hououin Kyouma (Sep 07 2025 at 11:24):
Eric Wieser said:
I don't know what you mean by "this" (the image in that message is misformatted).
I have posted this theorem in the topic.
Eric Wieser (Sep 07 2025 at 11:44):
I meant which is the one you said you couldn't find? (The first "this", not the second, sorry)
Eric Wieser (Sep 07 2025 at 11:44):
The one you pasted there is the one I pointed you to above
Hououin Kyouma (Sep 07 2025 at 11:51):
Eric Wieser said:
I meant which is the one you said you couldn't find? (The first "this", not the second, sorry)
The second theorem. And you have already told me that this is the theorem I need. Thanks a lot, I'll try again.
Bhavik Mehta (Sep 13 2025 at 08:10):
Eric Wieser said:
We have this only for at the moment
We have this for as well, see https://leanprover-community.github.io/mathlib4_docs/Mathlib/Analysis/CStarAlgebra/Matrix.html#Matrix.instL2OpNormedRing
Last updated: Dec 20 2025 at 21:32 UTC