Positivity and numerator growth for tropical cluster recurrences

Let (dn)(d_n) be a sequence satisfying

dn+N+dn=max(j=1N1[b1,j+1]+dn+j,j=1N1[b1,j+1]+dn+j),d_{n+N}+d_n=\max\left(\sum_{j=1}^{N-1}[b_{1,j+1}]_+d_{n+j},\sum_{j=1}^{N-1}[-b_{1,j+1}]_+d_{n+j}\right),

with initial conditions d1=1d_1=-1 and d2==dN=0d_2=\cdots=d_N=0 up to shifting the index, and suppose that (dn)(d_n) is not periodic. Let Nn\operatorname{N}_n denote the numerator in the associated Laurent representation. Tropical growth conjecture. Then (a) dn>0d_n>0 for all n>Nn>N, and (b) there is a constant C~>0\widetilde{\mathrm{C}}>0 such that

degNnC~dn(n).\deg\operatorname{N}_n\sim\widetilde{\mathrm{C}}d_n\qquad(n\to\infty).

These properties would relate denominator-vector growth to numerator degree growth and support the entropy analysis of the associated birational maps; the conjecture is not resolved in the supplied source.

Sources & referencesView supporting material

Primary source

Allan Fordy and Andrew Hone, “Discrete integrable systems and Poisson algebras from cluster maps”, arXiv:1207.6072 (2012).

Progress summary

Never refreshed

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

Solutions 0

No solutions have been posted yet.