Faltings's theorem
StructurePlace#✓fully proved
Let \(F\) be a field and \(K\) a field with \(F\) a \(K\)-algebra. A place of \(F/K\) is a valuation ring of \(F\) containing \(K\) and different from \(F\): a \(K\)-subalgebra \(\mathcal{O}\) of \(F\) with \(\mathcal{O} \neq F\) such that \(z \in \mathcal{O}\) or \(z^{-1} \in \mathcal{O}\) for every \(z \in F\). See [Stichtenoth, Definition 1.1.4 and 1.1.8].
Stichtenoth calls the maximal ideal of \(\mathcal{O}\) the place, which determines \(\mathcal{O}\), and also requires \(\mathcal{O} \neq K\); that condition is omitted since it is automatic.
@[ext] structure Place extends ValuationSubalgebra K F where
ne_top : toSubalgebra ≠ ⊤
@[ext] structure ValuationSubalgebra extends Subalgebra R F, ValuationSubring F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:63
TheoremFinitely many poles#●ready to prove
Let \(F\) be a function field over a field \(K\), that is, a finite extension of the rational function field \(K(t)\). For every \(x \in F\), there are only finitely many places \(P\) of \(F/K\) whose valuation ring \(\mathcal{O}_P\) does not contain \(x\), that is, \(x\) has only finitely many poles. See [Stichtenoth, Corollary 1.3.4].
public theorem_wanted finite_setOf_place_notMem [BundledFunctionField K F] (x : F) :
{v : Place K F | x ∉ v}.Finite
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
@[ext] structure Place extends ValuationSubalgebra K F where
ne_top : toSubalgebra ≠ ⊤
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:74
Sources
-. Problem faltings of the lean-eval benchmark.
Uses 1
AbbreviationAdele ring#✓fully proved
Let \(F\) be a field and \(K\) a field with \(F\) a \(K\)-algebra. The adele ring \(\mathcal{A}_F\) of \(F/K\) is the restricted product of copies of \(F\) indexed by the places \(P\) of \(F/K\), with respect to their valuation rings \(\mathcal{O}_P\) and the cofinite filter: its elements are the families \(\alpha = (\alpha_P)_P\) with \(\alpha_P \in F\) for every place \(P\) and \(\alpha_P \in \mathcal{O}_P\) for all but finitely many \(P\). See [Stichtenoth, Definition 1.5.2].
abbrev Adele : Type _ :=
RestrictedProduct (fun _ ↦ F) (fun v : Place K F ↦ (v : Set F)) Filter.cofinite
@[ext] structure Place extends ValuationSubalgebra K F where
ne_top : toSubalgebra ≠ ⊤
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:74
Uses 1 · Used by 5
DefinitionPrincipal adele map#●ready to state
Let \(F\) be a function field over a field \(K\), that is, a finite extension of the rational function field \(K(t)\). The principal adele map is a ring homomorphism from \(F\) to the adele ring \(\mathcal{A}_F\) of \(F/K\). It is meant to be the diagonal embedding sending \(x \in F\) to the adele whose coordinate at every place \(P\) of \(F/K\) is \(x\), which is well defined since \(x\) has only finitely many poles; this coordinate formula is stated separately.
public def_wanted Adele.principal [BundledFunctionField K F] : F →+* Adele K F
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
abbrev Adele : Type _ :=
RestrictedProduct (fun _ ↦ F) (fun v : Place K F ↦ (v : Set F)) Filter.cofinite
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:113
Sources
-. Problem faltings of the lean-eval benchmark.
Uses 2 · Used by 2
TheoremCoordinates of a principal adele#●statement formalized
Let \(F\) be a function field over a field \(K\), that is, a finite extension of the rational function field \(K(t)\). For every \(x \in F\) and every place \(P\) of \(F/K\), the coordinate at \(P\) of the principal adele of \(x\) is \(x\).
public theorem_wanted Adele.principal_apply [BundledFunctionField K F] (x : F)
(v : Place K F) : (❰Adele.principal❱ K x) v = x
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
@[ext] structure Place extends ValuationSubalgebra K F where
ne_top : toSubalgebra ≠ ⊤
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:74
public def_wanted Adele.principal [BundledFunctionField K F] : F →+* Adele K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:127
Sources
-. Problem faltings of the lean-eval benchmark.
Uses 3
TheoremThe adele algebra map is the principal embedding#●statement formalized
Let \(F\) be a function field over a field \(K\), that is, a finite extension of the rational function field \(K(t)\). The algebra map \(F \to \mathcal{A}_F\) of the \(F\)-algebra structure on the adele ring \(\mathcal{A}_F\) equals the principal adele map \(F \to \mathcal{A}_F\).
public theorem_wanted instAlgebraAdele_algebraMap [BundledFunctionField K F] :
algebraMap F (Adele K F) = ❰Adele.principal❱ K
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
abbrev Adele : Type _ :=
RestrictedProduct (fun _ ↦ F) (fun v : Place K F ↦ (v : Set F)) Filter.cofinite
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:113
public def_wanted Adele.principal [BundledFunctionField K F] : F →+* Adele K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:127
Sources
-. Problem faltings of the lean-eval benchmark.
Uses 3
DefinitionIntegral adeles#✓fully proved
Let \(F\) be a field and \(K\) a field with \(F\) a \(K\)-algebra. The integral adeles \(\mathcal{A}_F(0)\) form the \(K\)-subspace of the adele ring \(\mathcal{A}_F\) of \(F/K\) consisting of the adeles \(\alpha = (\alpha_P)_P\) whose coordinate \(\alpha_P\) lies in the valuation ring \(\mathcal{O}_P\) of \(P\) for every place \(P\) of \(F/K\). See [Stichtenoth, Definition 1.5.3].
def integralAdele : Submodule K (Adele K F) where
carrier := {x | ∀ v, x v ∈ v}
add_mem' h₁ h₂ v := add_mem (h₁ v) (h₂ v)
zero_mem' _ := zero_mem _
smul_mem' _ _ h v := v.smul_mem (h v) _
abbrev Adele : Type _ :=
RestrictedProduct (fun _ ↦ F) (fun v : Place K F ↦ (v : Set F)) Filter.cofinite
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:113
Uses 2 · Used by 1
DefinitionGenus of a function field#●ready to state
For a function field \(F\) over a field \(K\), that is, a finite extension of the rational function field \(K(t)\), the genus of \(F/K\) is a natural number \(g\), specified by the separately stated formula \(g = \dim_K \mathcal{A}_F / (\mathcal{A}_F(0) + F)\), where \(\mathcal{A}_F\) is the adele ring and \(\mathcal{A}_F(0)\) the integral adeles. See [Stichtenoth, Corollary 1.5.5].
public noncomputable def_wanted genus [BundledFunctionField K F] : ℕ
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
Sources
-. Problem faltings of the lean-eval benchmark.
TheoremGenus as the dimension of the adele class space#●statement formalized
Let \(F\) be a function field over a field \(K\), that is, a finite extension of the rational function field \(K(t)\). The genus \(g\) of \(F/K\) equals the dimension over \(K\) of the quotient \(\mathcal{A}_F / (\mathcal{A}_F(0) + F)\), where \(\mathcal{A}_F\) is the adele ring, \(\mathcal{A}_F(0)\) the integral adeles, and \(F\) is embedded in \(\mathcal{A}_F\) by the algebra map of the \(F\)-algebra structure on the adeles (stated separately to be the principal embedding). See [Stichtenoth, Corollary 1.5.5].
The dimension is the finite rank, taken to be \(0\) if the quotient is infinite-dimensional.
public theorem_wanted genus_eq [BundledFunctionField K F] :
❰genus❱ = Module.finrank K
(Adele K F ⧸ integralAdele K F ⊔ (IsScalarTower.toAlgHom K F _).toLinearMap.range)
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
public noncomputable def_wanted genus [BundledFunctionField K F] : ℕ
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:175
abbrev Adele : Type _ :=
RestrictedProduct (fun _ ↦ F) (fun v : Place K F ↦ (v : Set F)) Filter.cofinite
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:113
def integralAdele : Submodule K (Adele K F) where
carrier := {x | ∀ v, x v ∈ v}
add_mem' h₁ h₂ v := add_mem (h₁ v) (h₂ v)
zero_mem' _ := zero_mem _
smul_mem' _ _ h v := v.smul_mem (h v) _
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:164
Sources
-. Problem faltings of the lean-eval benchmark.
Uses 4
TheoremFaltings' theorem (Mordell conjecture)#●statement formalized
Let \(K\) be a number field and \(F\) a function field over \(K\), that is, a finite extension of the rational function field \(K(t)\). If the genus of \(F/K\) is at least \(2\), then \(F/K\) has only finitely many places of degree one, that is, places \(P\) whose residue field (the quotient of the valuation ring of \(P\) by its maximal ideal) has dimension \(1\) over \(K\).
public theorem_wanted faltings [NumberField K] [BundledFunctionField K F] :
2 ≤ ❰genus❱ K F →
{v : Place K F | Module.rank K (IsLocalRing.ResidueField v) = 1}.Finite
class BundledFunctionField extends
Algebra (RatFunc K) F, IsScalarTower K (RatFunc K) F, FunctionField K F
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:69
public noncomputable def_wanted genus [BundledFunctionField K F] : ℕ
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:175
@[ext] structure Place extends ValuationSubalgebra K F where
ne_top : toSubalgebra ≠ ⊤
WantedExt/AlgebraicGeometry/FaltingsWanted.lean:74
Sources
-. Problem faltings of the lean-eval benchmark.
Uses 2
Loading dependency explorer…
The dependency explorer needs JavaScript and its graph data.