Minimal-period conjecture for coefficients of restricted-partition quasi-polynomials

From papers

Let a1,a2,,ana_1,a_2,\ldots,a_n be positive integers, and let

r(x)=k0q(k)xk=1(1xa1)(1xa2)(1xan),r(x)=\sum_{k\geq 0}q(k)x^k=\frac{1}{(1-x^{a_1})(1-x^{a_2})\cdots(1-x^{a_n})},

where q(k)=cd(k)kd++c0(k)q(k)=c_d(k)k^d+\dots+c_0(k) is a quasi-polynomial. For a positive integer mm, define bm=#{i:mai}b_m=\#\{i:m\mid a_i\}. For 0jd0\leq j\leq d, define

pj=\lcm{m:bm>j}.p_j=\operatorname*{\lcm}\{m:b_m>j\}.

Minimal-period conjecture. The minimum period of cj(k)c_j(k) is pjp_j for every 0jd0\leq j\leq d. This conjecture predicts that the divisibility data of the aia_i determine the exact periods of all coefficient functions in this class of generating functions. The preceding discussion presents it as a more general statement that would imply the conjectured extension of the quasi-polynomial convolution theorem; its resolution is not given in the source.

Progress summary

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Sources & referencesView supporting material

Primary source

Matthias Beck, Steven Sam and Kevin Woods, “Maximal Periods of (Ehrhart) Quasi-Polynomials”, arXiv:math/0702242 (2007).

Solutions 0

No solutions have been posted yet.