Keyboard shortcuts

Press ← or → to navigate between chapters

Press ? to show this help

Press Esc to hide this help

Hanna’s Double-Composition Modulo-Four Classification

Abstract

The normalized series of OEIS A372577 has a complete modulo-four classification.

The entry cited in hanna2024a372577 defines A(x) by A(x)^2=A(A(xA(x)+xA(x)^2)), with constant coefficient zero and linear coefficient one. It conjectures remainder three at indices 6m-3 and remainder one at indices 6m-k for k in {0,1,2,4,5}, where m is positive. These indices cover all positive integers.

PowerSeries(Z) denotes formal power series over the integers, X is the indeterminate, coeff extracts a coefficient, and mk constructs a series from its coefficient function. The notation subst(f,u) means f(u), with the outer series first. All indices are natural numbers. The remainder of n is natural remainder, and the remainder of a(n) is integer remainder. The displayed approximation, argument, nested, and step operations are local notation for the private construction.

Definition 1.1 (Construction by stabilized coefficients).

Formalization. D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.generatingSeries (✓ std3).

Source. Repository-derived.

Commentary.

Writing A=XU and cancelling X^2U gives U=step(U). Each substitution argument is divisible by X^2. If two unit series agree below degree d, where d is positive, their transformed series agree below degree d+1. The approximations start at one, keep constant coefficient one, and stabilize through degree d by approximation d. Their diagonal coefficients define U and hence generatingSeries.

Definition 1.2 (The integer coefficient sequence).

Formalization. D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.a (✓ std3).

Source. Repository-derived.

Commentary.

The coefficient sequence is extracted from the constructed generating series.

Theorem 1.3 (The double composition and normalization).

Proof. Machine-checked in Lean as D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.generating_equation (✓ std3). ∎

Source. Repository-derived.

Commentary.

Coefficient stabilization proves that U is fixed by step. Multiplying this fixed-point identity by X^2*U restores the two nested substitutions in the defining equation. The constant and linear coefficients follow from the factor X and the constant coefficient one of U.

Theorem 1.4 (Uniqueness of the normalized solution).

Proof. Machine-checked in Lean as D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.generating_unique (✓ std3). ∎

Source. Repository-derived.

Commentary.

Factor any competing solution as X*U. The linear normalization makes U a unit, so cancelling X^2 and U in the equation gives the same fixed-point identity. Degree contraction proves equality of all coefficients.

Theorem 1.5 (Every positive coefficient modulo four).

Proof. Machine-checked in Lean as D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.hanna_conjecture (✓ std3). ∎

Resolves. Problems/oeis-a372577-double-composition-mod-four (proved) by D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.hanna_conjecture.

Citation. Paul D. Hanna (2024). OEIS A372577, expansion of g.f. A(x) satisfying A(x)^2 = A(A(xA(x) + xA(x)^2)). URL: https://oeis.org/A372577.

Commentary.

Over ZMod(4), put P=X+X^2+3X^3+X^4+X^5+X^6 and D=1-X^6. The rational series P/D satisfies the full double-composition equation: homogenizing P and D clears each substitution denominator, and two polynomial identities establish the result by unit cancellation. Uniqueness identifies it with the reduction of generatingSeries. The identity P/D=X/(1-X)+2X^3/(1-X^6), with formal unit inverses, then gives remainder three precisely at n mod 6 equal to three, and remainder one otherwise. This is both conjectured clauses together.

Theorem 1.6 (Oddness as a corollary).

Proof. Machine-checked in Lean as D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.odd_coefficients (✓ std3). ∎

Source. Repository-derived.

Commentary.

The complete classification gives only the residues one and three modulo four. Both imply integer remainder one modulo two, so every positive-index coefficient is odd.

References

  • Truth anchor: D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.a
  • Truth anchor: D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.generatingSeries
  • Truth anchor: D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.generating_equation
  • Truth anchor: D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.generating_unique
  • Truth anchor: D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.hanna_conjecture
  • Truth anchor: D5/S1/Recurrence/Residue/DoubleCompositionModFourClassification.odd_coefficients