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 : ), 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 : ), 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 : ), 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 : ), 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 #

@[implicit_reducible]

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

Equations
@[implicit_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 ) (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 ) {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 or .

    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 : ιAbsoluteValue R S} (h : ∀ (i : ι), (v i).IsNontrivial) (hv : Pairwise fun (i j : ι) => ¬(v i).IsEquiv (v j)) {ε : S} ( : 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 : ιAbsoluteValue R S} (h : ∀ (i : ι), (v i).IsNontrivial) (hv : Pairwise fun (i j : ι) => ¬(v i).IsEquiv (v j)) (a : ιR) {ε : S} ( : 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 ) :
    ∃ (w : AbsoluteValue K ), 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 } (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).