Zulip Chat Archive

Stream: Is there code for X?

Topic: sInf is minimum


Ira Fesefeldt (Jun 02 2024 at 14:14):

I am searching for a proof of something on the line

theorem sInf_of_finite {α : Type} [CompleteLinearOrder α] {s : Set α} (h_fin : Set.Finite s) (i : α) (P : α  Prop) :
    sInf {i  s | P i }  {i  s | P i } := sorry

That is, the infimum on a finite set with a linear order is the minimum of the finite set.

Does something like this exist already?

Ira Fesefeldt (Jun 02 2024 at 14:16):

Nevermind, already found... Set.Nonempty.csInf_mem

Kevin Buzzard (Jun 02 2024 at 14:17):

Your theorem as it stands is false BTW:

import Mathlib

theorem sInf_of_finite {α : Type} [CompleteLinearOrder α] {s : Set α} (h_fin : Set.Finite s) (i : α) (P : α  Prop) :
    sInf {i  s | P i }  {i  s | P i } := sorry

example : False := by
  have := sInf_of_finite (α := Fin 1) (s := ) (by simp) 0 (fun _  False)
  aesop

Ira Fesefeldt (Jun 02 2024 at 14:27):

Yes, nonempty is missing


Last updated: May 02 2025 at 03:31 UTC