Numerical Range convex
LinearPMap.numericalRange_convex
Project documentation
The Toeplitz-Hausdorff theorem.
Exact Lean statement
theorem numericalRange_convex (T : H →ₗ.[ℂ] H) : Convex ℝ (Θ T)
Formal artifact
Lean source
theorem numericalRange_convex (T : H →ₗ.[ℂ] H) : Convex ℝ (Θ T) := by intro z₀ hz₀ z₁ hz₁ a b ha hb hab rcases eq_or_ne z₁ z₀ with rfl | hz · simp [← add_mul, eq_sub_iff_add_eq.mpr hab, hz₁] · apply sub_ne_zero.mpr at hz obtain ⟨x₀, hx₀, _⟩ := hz₀ obtain ⟨x₁, hx₁, _⟩ := hz₁ -- Apply an affine transformation to effectively move the endpoints `z₀` and `z₁` to `0` and `1` let S : H →ₗ.[ℂ] H := (z₁ - z₀)⁻¹ • (T - z₀ • 1) let y₀ : S.domain := ⟨x₀, by simp [S, sub_domain]⟩ let y₁ : S.domain := ⟨x₁, by simp [S, sub_domain]⟩ have hy₀ : ‖y₀‖ = 1 := hx₀ have hy₁ : ‖y₁‖ = 1 := hx₁ have h₀ : ⟪↑y₀, S y₀⟫_ℂ = 0 := by simp_all [S, y₀, sub_apply, inner_smul_right, inner_sub_right] have h₁ : ⟪↑y₁, S y₁⟫_ℂ = 1 := by simp_all [S, y₁, sub_apply, inner_smul_right, inner_sub_right] suffices ofReal '' unitInterval ⊆ Θ S by have hba : a = 1 - b := by linarith rw [numericalRange_smul, numericalRange_sub_const] at this obtain ⟨c, ⟨d, ⟨x, hx, hxd⟩, hdc⟩, hca⟩ := (image_subset_iff.mp this) ⟨hb, by linarith⟩ simp only [mem_singleton_iff, exists_eq_left] at hca hdc hxd simp only [real_smul, smul_eq_mul] at * use x, hx simp_rw [hxd, hba, ofReal_sub, ofReal_one, ← hca, ← hdc] field_simp simp [mul_sub, mul_comm] -- First pick `θ` so that `y₂ ≔ eⁱᶿy₁` satisfies `⟪y₀, S y₂⟫ + ⟪y₂, S y₀⟫ ∈ ℝ` obtain ⟨θ, hθ⟩ := exists_phase_add_im_eq_zero ⟪↑y₀, S y₁⟫_ℂ ⟪↑y₁, S y₀⟫_ℂ let y₂ : S.domain := exp (I * θ) • y₁ have hy₂ : ‖y₂‖ = 1 := by simp [y₂, norm_smul, hy₁] have hy_im : (⟪↑y₀, S y₂⟫_ℂ).im = -(⟪↑y₂, S y₀⟫_ℂ).im := by apply eq_neg_iff_add_eq_zero.mpr simp [← hθ, y₂, map_smul, SetLike.val_smul, inner_smul_left, inner_smul_right, ← exp_conj] have h₂ : ⟪↑y₂, S y₂⟫_ℂ = 1 := by simp [y₂, map_smul, inner_smul_left, inner_smul_right, h₁, ← exp_conj, ← exp_add] -- `f` parametrizes the line connecting `y₀` and `y₂` and never vanishes because `y₀` and `y₂` -- are linearly independent (since `y₀ = λy₂` implies `0 = ⟪y₀, S y₀⟫ = |λ|²⟪y₂, S y₂⟫ = |λ|²`). let f : ℝ → S.domain := fun r ↦ (1 - r : ℂ) • y₀ + (r : ℂ) • y₂ have hf : ∀ r, f r ≠ 0 := by intro r hr rcases eq_or_ne r 0 with rfl | hr' · exact (hy₁ ▸ zero_ne_one).symm (norm_eq_zero.mpr (by simp_all [f])) · apply (pow_ne_zero 2 hr').symm apply ofReal_inj.mp calc _ = (1 - r) ^ 2 * ⟪↑y₀, S y₀⟫_ℂ := by simp [h₀] _ = ⟪↑((1 - r : ℂ) • y₀), S ((1 - r : ℂ) • y₀)⟫_ℂ := by simp [map_smul, inner_smul_left, inner_smul_right, ← mul_assoc, pow_two] _ = ⟪↑(-(r • y₂)), S (-(r • y₂))⟫_ℂ := by simp [eq_neg_iff_add_eq_zero.mpr hr] simp [map_neg, ← Complex.coe_smul, map_smul, inner_smul_left, inner_smul_right, pow_two, h₂] -- `g r = ⟪f r, S (f r)⟫_ℂ / ‖f r‖²` is real (by def of `θ`) and clearly in `Θ S`. -- `g 0 = 0`, `g 1 = 1` and continuity ensure that all of `[0,1]` is also in `Θ S`. let g : ℝ → ℝ := fun t ↦ (t ^ 2 + (1 - t) * t * (⟪↑y₀, S y₂⟫_ℂ + ⟪↑y₂, S y₀⟫_ℂ).re) / ‖f t‖ ^ 2 have hg₀ : g 0 = 0 := by simp [g] have hg₁ : g 1 = 1 := by simp [g, f, coe_norm y₂ ▸ hy₂] have hg_cont : Continuous g := Continuous.div₀ (by fun_prop) (by fun_prop) (by simp [hf]) intro c ⟨t, ht, htc⟩ obtain ⟨r, hr, hrt⟩ := (hg₀ ▸ hg₁ ▸ intermediate_value_Icc zero_le_one hg_cont.continuousOn) ht rw [← htc, ← hrt] refine ⟨‖f r‖⁻¹ • f r, ?_, ?_⟩ · simp only [mem_setOf_eq, norm_smul, norm_inv, norm_norm] exact inv_mul_cancel₀ (norm_ne_zero_iff.mpr (hf r)) · have hf_sq : ofReal (‖f r‖ ^ 2) ≠ 0 := by simp [hf] simp_rw [← Complex.coe_smul, map_smul, SetLike.val_smul, inner_smul_left,inner_smul_right, ← mul_assoc, conj_ofReal, ← pow_two, ← ofReal_pow, inv_pow, ofReal_inv] apply (inv_mul_eq_iff_eq_mul₀ hf_sq).mpr simp_rw [g, ofReal_div, mul_div_cancel₀ _ hf_sq, add_comm (r ^ 2)] simp only [f, map_add, map_smul, coe_add, inner_add_left, inner_add_right, SetLike.val_smul, inner_smul_left, inner_smul_right, h₀, h₂] nth_rw 1 [← re_add_im ⟪↑y₀, S y₂⟫_ℂ, ← re_add_im ⟪↑y₂, S y₀⟫_ℂ] simp only [hy_im, mul_add, RingHom.map_sub, RingHom.map_one, conj_ofReal, mul_zero, zero_add, ofReal_neg, neg_mul, mul_neg, mul_one, add_re, ofReal_add, ofReal_mul, ofReal_sub, ofReal_one, ofReal_pow] ring- Project
- Physlib
- License
- Apache-2.0
- Commit
- dd43e9e65791
- Source
- Physlib/QuantumMechanics/Operators/SpectralTheory/Basic.lean:543-615
Reuse this declaration
Bring the exact result into your workflow
The import identifies the source module. Your project still needs the pinned package dependency shown on this page.
What this badge means
This completion status comes from the project or community source. It has not yet been represented here as an independent rebuild and axiom audit.
Continue in this project
Related declarations
Adiabatic relation log
adiabatic_relation_log
Plain-language statement
Adiabatic relation in logarithmic form: If S(Ua,Va,N) = S(Ub,Vb,N) with N fixed, then c * log (Ua/Ub) + log (Va/Vb) = 0.
Source project: Physlib
Person-level attribution pending.
Adiabatic relation Ua Ub Va Vb
adiabatic_relation_UaUbVaVb
Plain-language statement
Adiabatic relation in product form: If S(Ua,Va,N) = S(Ub,Vb,N) with N fixed, then (Ua/Ub)^c * (Va/Vb) = 1.
Source project: Physlib
Person-level attribution pending.
Deriv Within mean Energy Beta eq neg variance
CanonicalEnsemble.derivWithin_meanEnergy_Beta_eq_neg_variance
Plain-language statement
(∂U/∂β) = -Var(E) for finite systems.
Source project: Physlib
Person-level attribution pending.