Documentation

TauCeti.AlgebraicGeometry.EllipticCurve.MinimalModel.ObstructionExponent

Local obstruction exponents of Weierstrass equations #

Let O be a Dedekind domain with fraction field K, let v be a height-one prime of O, and let W be an elliptic Weierstrass equation over K. The local obstruction exponent is

f_v(W) = (v(Delta W) - v(Delta_min,v)) / 12.

The numerator is divisible by twelve because an admissible change of variables scales the discriminant by the inverse twelfth power of its u-parameter. It is useful to define the exponent for every rational equation, with values in ℤ: local integrality is a sufficient hypothesis for nonnegativity, while a nonintegral equation may have negative defect.

These exponents are the local data assembled by the defect ideal of an integral equation. The reconstruction formula in this file is the interface that construction needs; it avoids relying on integer division or unfolding the definition.

Main definitions #

Main results #

References #

The difference between the discriminant valuation of an equation and the local minimal valuation is divisible by twelve. An admissible change of variables scales the discriminant by the inverse twelfth power of its u-parameter.

The local obstruction exponent f_v(W) = (v(Delta W) - v(Delta_min,v)) / 12.

It is integer-valued for every rational equation. Local integrality is not part of the definition: it is the hypothesis that makes the exponent nonnegative, as proved by obstructionExponentAt_nonneg_of_isIntegral.

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

    Twelve times the local obstruction exponent is the discriminant defect. This is the characteristic elimination lemma for obstructionExponentAt; consumers need not reason about integer division.

    @[simp]

    Changing variables subtracts the order of the scaling parameter from the local obstruction exponent.

    Local integrality makes the local obstruction exponent nonnegative. Without integrality the exponent remains defined but may be negative.

    @[simp]

    For a locally integral equation, the obstruction exponent vanishes exactly when the equation is minimal at that prime.

    An integral equation is globally minimal exactly when all of its local obstruction exponents vanish.

    A sharply semi-global model at v₀ is integral at v₀, minimal at every other height-one prime, and has discriminant defect exactly twelve at v₀. Unlike weak semi-global minimality, which permits any defect at its exceptional prime, this pins the defect to the smallest positive value, so the model's defect ideal is exactly 𝔭_{v₀}.

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

      A sharply semi-global model is integral at its exceptional prime.

      A sharply semi-global model is minimal away from its exceptional prime.

      The obstruction exponent of a sharply semi-global model at its exceptional prime is one.

      An equation integral at v₀, minimal away from v₀, and with obstruction exponent one at v₀ is sharply semi-global there.

      Every sharply semi-global model is semi-global in the weak, consumer-facing sense.