Documentation

Iwasawalib.NumberTheory.AbsoluteValue.DecompositionSubgroup

Decomposition subgroup for a place (AbsoluteValue) #

(To be written)

References:

Lemmas should go mathlib #

theorem exists_zpow_btwn_pow_of_lt {K : Type u_1} [Semifield K] [LinearOrder K] [IsStrictOrderedRing K] [Archimedean K] [ExistsAddOfLE K] {a b c : K} (h : a < b) (hb₀ : 0 < b) (hc₀ : 0 < c) (hc₁ : c ≠ 1) :
∃ (m : ℕ) (n : ℤ), a ^ m < c ^ n ∧ c ^ n < b ^ m

If a < b, b, c are positive and c ≠ 1, then there are m : ℕ and n : ℤ such that the power c ^ n is strictly between a ^ m and b ^ m.

TODO: go mathlib and add exists_pow_btwn_pow_of_lt

theorem exists_one_btwn_pow_mul_zpow_of_lt {K : Type u_1} [Semifield K] [LinearOrder K] [IsStrictOrderedRing K] [Archimedean K] [ExistsAddOfLE K] {a b c : K} (h : a < b) (hb₀ : 0 < b) (hc₀ : 0 < c) (hc₁ : c ≠ 1) :
∃ (m : ℕ) (n : ℤ), a ^ m * c ^ n < 1 ∧ 1 < b ^ m * c ^ n

If a < b, b, c are positive and c ≠ 1, then there are m : ℕ and n : ℤ such that 1 is strictly between a ^ m * c ^ n and b ^ m * c ^ n.

TODO: go mathlib

theorem zpow_unbounded_of_ne_one {K : Type u_1} [Semifield K] [LinearOrder K] [IsStrictOrderedRing K] [Archimedean K] [ExistsAddOfLE K] (x : K) {y : K} (hy0 : 0 < y) (hy1 : y ≠ 1) :
∃ (n : ℤ), x < y ^ n

TODO: go mathlib

theorem exists_pow_mul_zpow_lt_and_lt_pow_mul_zpow_of_lt {K : Type u_1} [Semifield K] [LinearOrder K] [IsStrictOrderedRing K] [Archimedean K] [ExistsAddOfLE K] {a b c d : K} (e : K) (h : a < b) (hb₀ : 0 < b) (hc₀ : 0 < c) (hc₁ : c ≠ 1) (hd₀ : 0 < d) :
∃ (m : ℕ) (n : ℤ), a ^ m * c ^ n < d ∧ e < b ^ m * c ^ n

If a < b, b, c, d are positive and c ≠ 1, e is any element, then there are m : ℕ and n : ℤ such that a ^ m * c ^ n < d and e < b ^ m * c ^ n.

TODO: go mathlib

Action on absolute values #

@[instance_reducible]

Ring isomorphisms act on the left on the absolute values by σ • v := v ∘ σ⁻¹.

Equations
@[instance_reducible]
instance AbsoluteValue.instMulActionAlgEquiv (F : Type u_1) (K : Type u_2) (S : Type u_3) [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] :

Algebra isomorphisms act on the left on the absolute values by σ • v := v ∘ σ⁻¹.

Equations
theorem AbsoluteValue.ringEquiv_smul_def {K : Type u_2} {S : Type u_3} [Semiring K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃+* K) :
σ • v = v.comp ⋯
@[simp]
theorem AbsoluteValue.ringEquiv_smul_apply {K : Type u_2} {S : Type u_3} [Semiring K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃+* K) (x : K) :
(σ • v) x = v (σ.symm x)
theorem AbsoluteValue.algEquiv_smul_def {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃ₐ[F] K) :
σ • v = v.comp ⋯
@[simp]
theorem AbsoluteValue.algEquiv_smul_apply {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃ₐ[F] K) (x : K) :
(σ • v) x = v (σ.symm x)
theorem AbsoluteValue.coe_algEquiv_smul_eq {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃ₐ[F] K) :
σ.toRingEquiv • v = σ • v

Decomposition subgroup for a place #

def AbsoluteValue.decompositionSubgroup (F : Type u_1) {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) :

The decomposition subgroup Dᵥ(K/F) in Gal(K/F) for a place v of K consists of all σ such that v ∘ σ = v. See [cassels1967algebraic], Chapter VII, §1. This definition also works for infinite places.

Note: Here we use def but not abbrev because sometimes simp lemmas for MulAction.stabilizer are not desired (will introduce σ⁻¹).

Equations
Instances For
    theorem AbsoluteValue.mem_decompositionSubgroup_iff {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃ₐ[F] K) :
    theorem AbsoluteValue.mem_decompositionSubgroup_iff_forall_apply_eq {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) (σ : K ≃ₐ[F] K) :
    σ ∈ decompositionSubgroup F v ↔ ∀ (x : K), v (σ x) = v x
    theorem AbsoluteValue.apply_eq_of_mem_decompositionSubgroup {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) {σ : K ≃ₐ[F] K} (h : σ ∈ decompositionSubgroup F v) (x : K) :
    v (σ x) = v x
    theorem AbsoluteValue.image_setOf_eq_of_mem_decompositionSubgroup {F : Type u_1} {K : Type u_2} {S : Type u_3} [CommSemiring F] [Semiring K] [Algebra F K] [Semiring S] [PartialOrder S] (v : AbsoluteValue K S) {σ : K ≃ₐ[F] K} (h : σ ∈ decompositionSubgroup F v) (y : S) :
    ⇑σ '' {x : K | v x ≤ y} = {x : K | v x ≤ y}

    The elements in decomposition subgroup preserve the set {x | v x ≤ y} for all y.

    theorem AbsoluteValue.decompositionSubgroup_comp_ringEquiv_eq_comap {S : Type u_3} [Semiring S] [PartialOrder S] {F : Type u_4} {E : Type u_5} [CommSemiring F] [Semiring E] [Algebra F E] {M : Type u_6} {N : Type u_7} [CommSemiring M] [Semiring N] [Algebra M N] {f : F →+* M} {g : E ≃+* N} (v : AbsoluteValue N S) (hsurj : Function.Surjective ⇑f) (hcomp : (algebraMap M N).comp f = (↑g).comp (algebraMap F E)) :
    theorem AbsoluteValue.mem_decompositionSubgroup_iff_image_setOf_eq {F : Type u_1} {K : Type u_2} [Field F] [Field K] [Algebra F K] [Algebra.IsAlgebraic F K] (v : AbsoluteValue K ℝ) (σ : Gal(K/F)) :
    σ ∈ decompositionSubgroup F v ↔ ⇑σ '' {x : K | v x ≤ 1} = {x : K | v x ≤ 1}

    If K / F is algebraic, then the decomposition subgroup Dᵥ(K/F) is also the subgroup consists of all σ preserving the set {x | v x ≤ 1}.

    NOTE: This is not true if K / F is not algebraic, for example, when K is the fraction field of the valuation ring $\bigcup_{n=1}^\infty F⟦X^{1/n}⟧$, k ≥ 2 is a fixed positive integer, σ is $X^{1/n} \mapsto X^{k/n}$, then it preserves the valuation ring but doesn't preserve the valuation.

    theorem AbsoluteValue.continuous_of_mem_decompositionSubgroup {F : Type u_1} {K : Type u_2} [Field F] [Field K] [Algebra F K] (v : AbsoluteValue K ℝ) {σ : Gal(K/F)} (h : σ ∈ decompositionSubgroup F v) :

    The elements in decomposition subgroup are continuous in v-adic topology.

    If K / F is algebraic, then an element is contained in the decomposition subgroup of v if and only if it is continuous under the v-adic topology.

    If L / K / F is an extension tower with L / F Galois, v is a place of L, then Dᵥ(L/K) = Dᵥ(L/F) ⊓ Gal(L/K).

    If L / K / F is an extension tower with L / F Galois, v is a place of L, then Dᵥ(L/K) ≤ Dᵥ(L/F).

    Decomposition subgroup for an infinite place is either trivial or ℤ/2ℤ. This is because the decomposition subgroup can be viewed as Galois group of completions of fields with respect to valuations (didn't formalized yet), in the archimedean case they must be ℝ or ℂ.

    theorem AbsoluteValue.exists_sub_one_lt_and_lt_of_not_isEquiv {R : Type u_1} {S : Type u_2} [Field R] [Field S] [LinearOrder S] [TopologicalSpace S] [IsStrictOrderedRing S] [Archimedean S] [OrderTopology S] {ι : Type u_3} [Finite ι] {v : ι → AbsoluteValue R S} (h : ∀ (i : ι), (v i).IsNontrivial) (hv : Pairwise fun (i j : ι) => ¬(v i).IsEquiv (v j)) {ε : S} (hε : 0 < ε) (i : ι) :
    ∃ (z : R), (v i) (z - 1) < ε ∧ ∀ (j : ι), j ≠ i → (v j) z < ε

    If v : ι → AbsoluteValue R S is a finite collection of non-trivial and pairwise inequivalent absolute values, then for any ε > 0 and any i : ι there is some z : R such that v i (z - 1) < ε and v j z < ε for all j ≠ i.

    TODO: go mathlib

    theorem AbsoluteValue.exists_sub_lt_of_not_isEquiv {R : Type u_1} {S : Type u_2} [Field R] [Field S] [LinearOrder S] [TopologicalSpace S] [IsStrictOrderedRing S] [Archimedean S] [OrderTopology S] {ι : Type u_3} [Finite ι] {v : ι → AbsoluteValue R S} (h : ∀ (i : ι), (v i).IsNontrivial) (hv : Pairwise fun (i j : ι) => ¬(v i).IsEquiv (v j)) (a : ι → R) {ε : S} (hε : 0 < ε) :
    ∃ (x : R), ∀ (i : ι), (v i) (x - a i) < ε

    A version of Approximation Theorem: if v : ι → AbsoluteValue R S is a finite collection of non-trivial and pairwise inequivalent absolute values, a : ι → R is a sequence of elements in R, then for any ε > 0 there is some x : R such that v i (x - a i) < ε for all i. See [Neukirch1992], II.3.4.

    TODO: go mathlib

    theorem AbsoluteValue.exists_liesOver {F : Type u_1} (K : Type u_2) [Field F] [Field K] [Algebra F K] [Algebra.IsAlgebraic F K] (v : AbsoluteValue F ℝ) :
    ∃ (w : AbsoluteValue K ℝ), w.LiesOver v

    If K / F is an algebraic extension, then any place v of F can be extended to K. (Is this correct?)

    theorem AbsoluteValue.exists_algEquiv_comp_eq_of_comp_eq {F : Type u_1} {K : Type u_2} [Field F] [Field K] [Algebra F K] [Normal F K] {v w : AbsoluteValue K ℝ} (h : v.comp ⋯ = w.comp ⋯) :
    ∃ (σ : Gal(K/F)), v.comp ⋯ = w

    If K / F is a normal extension, then any two places of K which coincide when restrict to F are conjugate by an element of Gal(K/F). See [Neukirch1992], II.9.1.

    If L / K / F is an extension tower, L / F and K / F are Galois, v and w are places of K and L, respectively, such that w lies over v, then the image of D_w(L/F) in Gal(K/F) is equal to Dᵥ(K/F).