Documentation

TauCeti.NumberTheory.NumberField.Quadratic.Conjugation.Ambiguous.Basic

Ambiguous ideals of a quadratic field #

An ideal of 𝓞 K is ambiguous when quadratic conjugation σ fixes it, σI = I; an ideal class is ambiguous when σ fixes it, which for a quadratic field means exactly that the class is 2-torsion (NumberField.mulEquiv_ringOfIntegersQuadraticConj_apply_eq_self_iff, since σ acts by inversion). Every ambiguous ideal obviously has an ambiguous class. This file proves the converse for an imaginary quadratic field: every ambiguous class is the class of an ambiguous ideal, so the two notions of "ambiguous" match up:

C ^ 2 = 1 ↔ ∃ I, σI = I ∧ [I] = C.

This is the descent step of the classical ambiguous class number formula, which counts the 2-torsion of Cl(𝓞 K): it replaces a count of ambiguous classes by a count of ambiguous ideals. The remaining step of that count — that the class of an ambiguous ideal is a product of classes of ramified primes, which in turn satisfy the relation ∏ 𝔭 = (θ) of TauCeti.Multiquadratic.span_singleton_eq_prod_primeFactors — is NumberField.classGroupMk0_mem_closure_of_map_eq_self in Quadratic/Conjugation/Ambiguous/Ideal.lean; the two together give the genus-theoretic bound 2-rank ≤ t - 1 of the Multiquadratic roadmap.

The proof is Hilbert's Theorem 90 for the quadratic extension K/ℚ, in the elementary form available for a degree-two extension. If [σJ] = [J] then (x) σJ = (y) J for nonzero x y : 𝓞 K; conjugating and cancelling gives (x σx) = (y σy), so x σx and y σy are associates. The unit relating them has norm the square of the rational N(y) / N(x), and in a totally complex field total positivity is vacuous, so the norm of a nonzero element is strictly positive (NumberField.norm_pos_of_isTotallyPositive), which forces x σx = y σy on the nose. Hilbert 90 then produces ε ≠ 0 with x ε = y σε, and I = (ε) J is an ambiguous ideal in J's class. For a real quadratic field the positivity fails — that is exactly where the classical unit index [E : E ∩ N K^×] enters the ambiguous class number formula — so the hypothesis IsTotallyComplex K is essential rather than technical. The repair is to pass to the narrow class group, where total positivity restores the sign for every quadratic field; see Quadratic/Conjugation/Ambiguous/Narrow.lean.

The integral Hilbert 90 result used below is proved in Quadratic.Conjugation.Hilbert90.

For a quadratic field of either signature, this file also defines a strongly ambiguous class to be an ordinary ideal class represented by an ambiguous ideal. The definition and its basic API do not require the totally complex hypothesis; only the converse from an ambiguous class to an ambiguous ideal does.

See F. Lemmermeyer, Reciprocity Laws: From Euler to Eisenstein, §2.2, whose Proposition 2.9 is the statement proved here, and D. A. Cox, Primes of the Form x² + ny², §6.A, for the classical ambiguous class number formula.

Main results #

theorem Ideal.map_map_of_involutive {R : Type u_2} [CommSemiring R] {f : R ≃+* R} (hf : Function.Involutive ⇑f) (I : Ideal R) :
map f (map f I) = I

Pushing an ideal forward twice along an involutive ring automorphism returns it.

theorem NumberField.mul_ringOfIntegersQuadraticConj_eq_mul_ringOfIntegersQuadraticConj_of_associated {K : Type u_1} [Field K] [NumberField K] {θ : RingOfIntegers K} {d : ℤ} (hmin : minpoly ℤ θ = Polynomial.X ^ 2 - Polynomial.C d) (hgen : ℚ[↑θ] = ⊤) {x y : RingOfIntegers K} (hnorm : 0 < (Algebra.norm ℚ) (↑x * ↑y)) (hassoc : Associated (x * (ringOfIntegersQuadraticConj hmin hgen) x) (y * (ringOfIntegersQuadraticConj hmin hgen) y)) :

Two elements with associated norms and positive norm product have equal norms. If x σx and y σy are associates in 𝓞 K and the norm of x y is positive, then x σx = y σy: their ratio is a unit whose norm is the square of the rational N(y) / N(x), so N(x) = ± N(y), and the sign is fixed by 0 < N(x) N(y).

This is where the archimedean input enters the ambiguous class number formula. For a totally complex field the norm of a nonzero element is automatically positive; for a real quadratic field the positivity has to come from total positivity of x y, which is what the narrow class group supplies.

Every ambiguous ideal class of an imaginary quadratic field is the class of an ambiguous ideal. Let K be a totally complex quadratic number field with quadratic conjugation σ. A 2-torsion ideal class — equivalently, by mulEquiv_ringOfIntegersQuadraticConj_apply_eq_self_iff, a class fixed by σ — is the class of an ideal I with σI = I. This is the Hilbert-90 descent step of the ambiguous class number formula.

The class of an ambiguous ideal is 2-torsion. An ideal fixed by quadratic conjugation has a class fixed by the induced action on Cl(𝓞 K), which is 2-torsion because that action is inversion (mulEquiv_ringOfIntegersQuadraticConj_apply_eq_self_iff). This is the easy direction of sq_eq_one_iff_exists_map_ringOfIntegersQuadraticConj_eq_self, and needs no hypothesis on the signature of K.

def NumberField.IsStronglyAmbiguousClass {K : Type u_1} [Field K] [NumberField K] {θ : RingOfIntegers K} {d : ℤ} (hmin : minpoly ℤ θ = Polynomial.X ^ 2 - Polynomial.C d) (hgen : ℚ[↑θ] = ⊤) (C : ClassGroup (RingOfIntegers K)) :

An ordinary ideal class is strongly ambiguous if it is represented by an ideal fixed by quadratic conjugation.

Equations
  • One or more equations did not get rendered due to their size.
Instances For

    A class is strongly ambiguous exactly when it has a representative fixed by quadratic conjugation.

    theorem NumberField.IsStronglyAmbiguousClass.intro {K : Type u_1} [Field K] [NumberField K] {θ : RingOfIntegers K} {d : ℤ} (hmin : minpoly ℤ θ = Polynomial.X ^ 2 - Polynomial.C d) (hgen : ℚ[↑θ] = ⊤) (I : ↥(nonZeroDivisors (Ideal (RingOfIntegers K)))) (hI : Ideal.map (ringOfIntegersQuadraticConj hmin hgen) ↑I = ↑I) :

    A conjugation-fixed ideal represents a strongly ambiguous class.

    theorem NumberField.IsStronglyAmbiguousClass.elim {K : Type u_1} [Field K] [NumberField K] {x✝ : RingOfIntegers K} {a✝ : ℤ} {hmin : minpoly ℤ x✝ = Polynomial.X ^ 2 - Polynomial.C a✝} {hgen : ℚ[↑x✝] = ⊤} {C : ClassGroup (RingOfIntegers K)} (hC : IsStronglyAmbiguousClass hmin hgen C) :

    A strongly ambiguous class has a representative fixed by quadratic conjugation.

    Every strongly ambiguous class is fixed by quadratic conjugation.

    noncomputable def NumberField.stronglyAmbiguousClassSubgroup {K : Type u_1} [Field K] [NumberField K] {θ : RingOfIntegers K} {d : ℤ} (hmin : minpoly ℤ θ = Polynomial.X ^ 2 - Polynomial.C d) (hgen : ℚ[↑θ] = ⊤) :

    The subgroup of the ordinary class group consisting of strongly ambiguous classes.

    Equations
    • One or more equations did not get rendered due to their size.
    Instances For
      @[simp]
      theorem NumberField.mem_stronglyAmbiguousClassSubgroup {K : Type u_1} [Field K] [NumberField K] {x✝ : RingOfIntegers K} {a✝ : ℤ} {hmin : minpoly ℤ x✝ = Polynomial.X ^ 2 - Polynomial.C a✝} {hgen : ℚ[↑x✝] = ⊤} {C : ClassGroup (RingOfIntegers K)} :

      Membership in the strongly ambiguous class subgroup means having a representative fixed by quadratic conjugation.

      The ambiguous classes of an imaginary quadratic field are exactly the classes of ambiguous ideals. For a totally complex quadratic number field, an ideal class is 2-torsion — equivalently fixed by quadratic conjugation — precisely when it is represented by an ideal that conjugation fixes. This is the descent step of the ambiguous class number formula: it turns the count of 2-torsion classes into a count of ambiguous ideals. Classifying those ideals — their classes are products of classes of ramified primes — is classGroupMk0_mem_closure_of_map_eq_self.