Documentation

Complexity.Computability.Circuit.Rational.CyclotomicGrid

Cyclotomic fields and a common evaluation grid #

The rational cyclotomic field ℚ[X] / (X ^ (2 ^ e) + 1) of 2 ^ (e + 1)-th roots of unity, with its power basis, its primitive additive character on ZMod (2 ^ (e + 1)) and its regular representation by 2 ^ e × 2 ^ e rational matrices. Over any field, a Schwartz–Zippel argument shows that one fixed grid S ^ k simultaneously avoids the zeros of a whole list of nonzero polynomials of bounded total degree, provided S is large enough.

Main definitions #

Main results #

References #

Degree of the chosen cyclotomic modulus.

Instances For

    Order of the chosen root of unity.

    Instances For

      The order of the root of unity is nonzero.

      The cyclotomic modulus for roots of unity of power-of-two order.

      Instances For

        The modulus X ^ (2 ^ e) + 1 is the 2 ^ (e + 1)-th cyclotomic polynomial, hence irreducible over ℚ.

        @[reducible, inline]

        The rational cyclotomic quotient used for the matrix construction.

        Instances For

          The quotient's power basis generated by the adjoined root.

          Instances For

            Reindex the power basis by the explicit cyclotomic degree.

            Instances For

              The cyclotomic quotient contains ℚ, so it has characteristic zero.

              The cyclotomic quotient is finite-dimensional over ℚ.

              An embedding of the cyclotomic quotient into the complex numbers.

              Instances For

                The adjoined root is a primitive root of unity of order 2 ^ (e + 1). Since the modulus is irreducible, the quotient is a field, so invertibility in the regular representation can be tested at this single primitive root.

                The additive character generated by the primitive cyclotomic root.

                Instances For

                  The regular rational matrix representation of the cyclotomic field.

                  Instances For
                    theorem Cslib.Circuits.Rational.FiniteGrid.exists_eval_ne_zero {K : Type u_1} [Field K] {k : ℕ} (p : MvPolynomial (Fin k) K) (hp : p ≠ 0) (S : Finset K) (hS : p.totalDegree < S.card) :
                    ∃ (x : Fin k → K), (∀ (i : Fin k), x i ∈ S) ∧ (MvPolynomial.eval x) p ≠ 0

                    The fixed grid, independent of the coefficients, finds a nonzero value.

                    theorem Cslib.Circuits.Rational.FiniteGrid.det_totalDegree_le {K : Type u_1} [Field K] {k : ℕ} {ι : Type u_2} [Fintype ι] [DecidableEq ι] (A : Matrix ι ι (MvPolynomial (Fin k) K)) (d : ℕ) (hA : ∀ (i j : ι), (A i j).totalDegree ≤ d) :

                    A determinant of polynomials of total degree at most d has total degree at most |ι| * d.

                    theorem Cslib.Circuits.Rational.FiniteGrid.simultaneous {K : Type u_1} [Field K] {k : ℕ} (P : List (MvPolynomial (Fin k) K)) (hP : ∀ p ∈ P, p ≠ 0) (d : ℕ) (hd : ∀ p ∈ P, p.totalDegree ≤ d) (S : Finset K) (hS : P.length * d < S.card) :
                    ∃ (x : Fin k → K), (∀ (i : Fin k), x i ∈ S) ∧ ∀ p ∈ P, (MvPolynomial.eval x) p ≠ 0

                    The common grid specializes a list of determinant conditions simultaneously, rather than one formula gate at a time.