Normalized MacWilliams identities #
The weight enumerator of a dual code is the MacWilliams substitution divided by the
cardinality of the code. For a self-dual code over a field of size q, homogeneity absorbs
this factor into the variables: the enumerator is fixed by
(X, Y) ↦ ((X + (q - 1) Y) / √q, (X - Y) / √q).
The statements allow values in arbitrary algebras, so they apply to polynomial variables
as well as numerical evaluations. The coefficient normalization is over ℚ, whereas
the variable normalization only requires an invertible square root of q. The real-algebra
corollary uses the positive square root.
The rational identity also holds over a finite commutative ring carrying a primitive
additive character with values in a characteristic zero domain.
References #
F. J. MacWilliams and N. J. A. Sloane, The Theory of Error-Correcting Codes, North-Holland (1977), Chapter 5, §2; W. C. Huffman and V. Pless, Fundamentals of Error-Correcting Codes, Cambridge University Press (2003), §7.2.
The rational normalized MacWilliams identity over a finite commutative ring
carrying a primitive additive character into a characteristic zero domain, evaluated
in any commutative ℚ-algebra.
The rational normalized MacWilliams identity for a linear code over a finite
field F with q elements: W_{C⊥}(x, y) = (1/#C) · W_C(x + (q-1)y, x - y).
The variables may lie in any commutative ℚ-algebra, including ℚ[X, Y].
A self-dual code's weight enumerator is fixed by the MacWilliams substitution
normalized by any invertible square root s of the alphabet size, with inverse t.
This holds over any commutative ring containing such a square root.
A self-dual code's weight enumerator is fixed by the normalized MacWilliams
substitution. The variables may lie in any commutative real algebra; in particular
this is an identity in ℝ[X, Y].