Chapter LorentzRealRepDirect
DefinitionChapterLorentzRealRepDirectdefinitionsformalizationtimepiece
Formal definitions for the timepiece Lean 4 formalization (source chapter BookProof/ChapterLorentzRealRepDirect.lean): generated def bundle for ChapterLorentzRealRepDirect. See BookProof/ChapterLorentzRealRepDirect.lean for full context.
Definition code
import Definitions.Def_ChapterLorentzRealRep
import Definitions.Def_ChapterLorentzRealRepSum
import Definitions.Def_ChapterLorentzRealRepFull
import Definitions.Def_ChapterA3
import Definitions.Def_ChapterPinOmega
import Mathlib
/-!
# Chapter "Real representations, CPT theorem and the relativistic position operator",
§"On the Lorentz, SL(2,C) and Pin(3,1) groups", **Lemma 52** — the four
representation spaces form an *internal direct sum*
Continuing `BookProof.ChapterLorentzRealRep`, `ChapterLorentzRealRepSum` and
`ChapterLorentzRealRepFull` (`book.tex` Lemma 52, line ~5560), which built the
four mutually Frobenius-orthogonal real representation spaces
* `WHalf` — the `(1/2,1/2)` vector representation (dimension `4`),
* `W10` — the `(1,0)` representation (dimension `6`),
* `WPs` — the pseudo-`(1/2,1/2)` representation (dimension `4`),
* `WTwo` — the two discrete Majorana directions `span{iγ⁰, γ⁰γ⁵}` (dimension `2`),
and proved that their fourfold sup is *all* of `Matrix (Fin 4) (Fin 4) ℝ`
(`decomposition_top`) with `16 = 4 + 6 + 4 + 2` (`finrank_full_eq_add`).
Those earlier files certified the *internal direct sum* only through a dimension
count. This file upgrades that to the genuine structural statement:
`DirectSum.IsInternal` of the family `WFam = ![WHalf, W10, WPs, WTwo]`, i.e. the
canonical map `⨁ᵢ WFam i → Matrix (Fin 4) (Fin 4) ℝ` is an isomorphism. From it
we read off the two defining properties of an internal direct sum:
* **spanning** `iSup_WFam_eq_top` (`⨆ i, WFam i = ⊤`);
* **independence** `WFam_iSupIndep` (`iSupIndep WFam`);
and the headline `WFam_isInternal`.
The proof is the finite-dimensional bijectivity argument: the canonical map
`DirectSum.coeLinearMap WFam` is surjective (its range is `⨆ i, WFam i = ⊤`) and
its domain `⨁ᵢ WFam i` has dimension `∑ᵢ finrank (WFam i) = 16 = finrank of the
whole matrix algebra`, so a surjection between equidimensional finite spaces is a
bijection.
As requested this stays **off the gravity line** and **off the Hankel-transform
line** (only the Majorana representation spaces of the previous waves are used).
Everything is `sorry`-free and `axiom`-free.
-/
open Matrix Module
namespace BookProof.ChapterLorentzRealRepDirect
open BookProof.ChapterLorentzRealRep BookProof.ChapterLorentzRealRepSum
open BookProof.ChapterLorentzRealRepFull
/-- The family of the four real representation subspaces of `SL(2,ℂ)` inside the
`16`-dimensional matrix algebra. -/
noncomputable def WFam : Fin 4 → Submodule ℝ (Matrix (Fin 4) (Fin 4) ℝ) :=
![WHalf, W10, WPs, WTwo]
/-
The fourfold supremum of the family is the whole matrix algebra.
-/
/-
The dimensions of the four summands add up to the dimension of the whole
matrix algebra (`4 + 6 + 4 + 2 = 16`).
-/
/-
**Internal direct sum.** The four mutually orthogonal representation spaces
form an internal direct sum of the whole matrix algebra: the canonical map
`⨁ᵢ WFam i → Matrix (Fin 4) (Fin 4) ℝ` is an isomorphism.
-/
end BookProof.ChapterLorentzRealRepDirect
Source