The trivial ideal weight and Dedekind zeta coefficients #
This file identifies the norm coefficients of the trivial ideal weight with the coefficients of
the Dedekind zeta function. There is one necessary exception: Mathlib's coefficient counts all
integral ideals and therefore has value 1 at index zero, contributed by the zero ideal, whereas
an ArithmeticFunction has value zero there. Since LSeries ignores its zero coefficient, the
two coefficient systems define the same series.
Since 1 * 1 = 1 pointwise, the same identification shows that regrouping has no
pointwise-product formula as soon as some norm is attained by two ideals
(TauCeti.not_forall_normCoeff_mul_eq_pmul): the coefficient of the product is the ideal count,
not its square. ℚ(i), with its two ideals of norm 5, is such a field.
For the rational field the ring of integers is isomorphic to ℤ. Mapping an ideal through this
isomorphism and using Int.ideal_span_absNorm_eq_self shows that there is exactly one ideal of
each positive norm. Thus the trivial ideal weight over ℚ regroups to the constant coefficient
1 at every positive index, as for the Riemann zeta function.
Roadmap role #
This is Layer 1.3, the trivial specialization, of
TauCetiRoadmap/ArithmeticDirichletSeries/README.md. It completes Layer 1 without asserting the
exact abscissa of convergence; that is Layer 5, proved in
TauCeti.abscissaOfAbsConv_normCoeff_one.
The coefficient used in Mathlib's definition of the Dedekind zeta function: the number of
integral ideals of absolute norm n.
Unlike an ArithmeticFunction, this function has value 1 at zero, contributed by the zero
ideal.
Equations
- TauCeti.dedekindZetaCoeff K n = Nat.card { I : Ideal (NumberField.RingOfIntegers K) // Ideal.absNorm I = n }
Instances For
The zero ideal is the unique integral ideal of absolute norm zero.
The Dedekind zeta function is the LSeries of dedekindZetaCoeff.
Away from zero, the cardinality of the finite nonzero-ideal norm fibre is the corresponding Dedekind zeta coefficient.
The trivial ideal arithmetic function regroups to the Dedekind zeta coefficients away from
zero. At zero its norm coefficient is forced to vanish by the ArithmeticFunction carrier.
The trivial unitary ideal weight has the Dedekind zeta coefficients away from zero.
Regrouping the trivial ideal weight gives Mathlib's Dedekind zeta function.
Regrouping does not turn pointwise products into pointwise products. If two or more
integral ideals share the absolute norm n, then no formula
normCoeff (f * g) = (normCoeff f).pmul (normCoeff g) holds: already for f = g = 1 the
n-th coefficient of 1 * 1 = 1 is the ideal count dedekindZetaCoeff K n, not its square.
There is exactly one integral ideal of ℚ of each absolute norm.
Over ℚ, a nonzero integral ideal is the only one of its absolute norm.
Over ℚ, the trivial unitary ideal weight has coefficient 1 at every positive integer.