arXiv ScienceSearch

arXiv subjects

Eric Mortenson

Publications and source records attributed to Eric Mortenson.

10 recordsLinked to original sources

Dyson's Ranks and Appell-Lerch Sums

Denote by $p(n)$ the number of partitions of $n$ and by $N(a,M;n)$ the number of partitions of $n$ with rank congruent to $a$ modulo $M$. We find and prove a general formula for Dyson's ranks by considering the deviation of the ranks from the average: \begin{equation*} D(a,M) := \sum_{n= 0}^{\infty}\left(N(a,M;n) - \frac{p(n)}{M}\right) q^n. \end{equation*} Using Appell--Lerch sum properties we decompose $D(a,M)$ into modular and mock modular parts so that the mock modular component is supported on certain arithmetic progressions, whose modulus we can control. Using our decomposition, we show how our formula gives as a straightforward consequence Atkin and Swinnerton-Dyer's results on ranks as well as Bringmann, Ono, and Rhoades's results on Maass forms. We also apply our techniques to a variation of Dyson's ranks due to Berkovitch and Garvan.

math.NT

On the dual nature of partial theta functions and Appell-Lerch sums

In recent work, Hickerson and the author demonstrated that it is useful to think of Appell--Lerch sums as partial theta functions. This notion can be used to relate identities involving partial theta functions with identities involving Appell--Lerch sums. In this sense, Appell--Lerch sums and partial theta functions appear to be dual to each other. This duality theory is not unlike that found by Andrews between various sets of identities of Rogers-Ramanujan type with respect to Baxter's solution to the hard hexagon model of statistical mechanics. As an application we construct bilateral $q$-series with mixed mock modular behaviour.

math.NT

Ramanujan's radial limits and mixed mock modular bilateral $q$-hypergeometric series

Using results from Ramanujan's lost notebook, Zudilin recently gave an insightful proof of a radial limit result of Folsom, Ono, and Rhoades for mock theta functions. Here we see that the author's previous work on the dual nature of Appell--Lerch sums and partial theta functions and on constructing bilateral $q$-series with mixed mock modular behaviour is well-suited for such radial limits. We present five more radial limit results which follow from mixed mock modular bilateral $q$-hypergeometric series. We also obtain the mixed mock modular bilateral series for a universal mock theta function of Gordon and McIntosh. The later bilateral series can be used to compute radial limits for many classical second, sixth, eight, and tenth order mock theta functions.

math.NT

Eulerian series as modular forms revisited

Recently, Bringmann, Ono, and Rhoades employed harmonic weak Maass forms to prove results on Eulerian series as modular forms. By changing the setting to Appell--Lerch sums, we shorten the proof of one of their main theorems. In addition we discuss connections to recent work of Kang.

math.NT

On three third order mock theta functions and Hecke-type double sums

We obtain four Hecke-type double sums for three of Ramanujan's third order mock theta functions. We discuss how these four are related to the new mock theta functions of Andrews' work on $q$-orthogonal polynomials and Bringmann, Hikami, and Lovejoy's work on unified Witten-Reshetikhin-Turaev invariants of certain Seifert manifolds. We then prove identities between these new mock theta functions by first expressing them in terms of the universal mock theta function.

math.NT

Threefield identities and simultaneous representations of primes by binary quadratic forms

Kaplansky [2003] proved a theorem on the simultaneous representation of a prime $p$ by two different principal binary quadratic forms. Later, Brink found five more like theorems and claimed that there were no others. By putting Kaplansky-like theorems into the context of threefield identities after Andrews, Dyson, and Hickerson, we find that there are at least two similar results not on Brink's list. We also show how such theorems are related to results of Muskat on binary quadratic forms

math.NT

Hecke-type double sums, Appell-Lerch sums, and mock theta functions (I)

By developing a connection between partial theta functions and Appell-Lerch sums, we find and prove a formula which expresses Hecke-type double sums in terms of Appell-Lerch sums and theta functions. Not only does our formula prove classical Hecke-type double sum identities such as those found in work Kac and Peterson on affine Lie Algebras and Hecke modular forms, but once we have the Hecke-type forms for Ramanujan's mock theta functions our formula gives straightforward proofs of many of the classical mock theta function identities. In particular, we obtain a new proof of the mock theta conjectures. Our formula also applies to positive-level string functions associated with admissable representations of the affine Lie Algebra $A_1^{(1)}$ as introduced by Kac and Wakimoto.

math.NT

Chiral potentials, perturbation theory, and the 1S0 channel of NN scattering

We use nucleon-nucleon phase shifts obtained from experimental data, together with the chiral expansion for the long-distance part of the NN interaction, to obtain information about the short-distance piece of the NN potential that is at work in the 1S0 channel. We find that if the scale R that defines the separation between "long-" and "short-" distance is chosen to be \lsim 1.8 fm then the energy dependence produced by short-distance dynamics is well approximated by a two-term polynomial for Tlab < 200 MeV. We also find that a quantitative description of NN dynamics is possible, at least in this channel, if one treats the long-distance parts of the chiral NN potential in perturbation theory. However, in order to achieve this we have to choose a separation scale R that is larger than 1.0 fm.

nucl-th

Number theoretic properties of Wronskians of Andrews-Gordon series

For positive integers $1\leq i\leq k$, we consider the arithmetic properties of quotients of Wronskians in certain normalizations of the Andrews-Gordon $q$-series $$ \prod_{1\leq n\not \equiv 0,\pm i\pmod{2k+1}}\frac{1}{1-q^n}. $$ This study is motivated by their appearance in conformal field theory, where these series are essentially the irreducible characters of $(2,2k+1)$ Virasoro minimal models. We determine the vanishing of such Wronskians, a result whose proof reveals many partition identities. For example, if $P_{b}(a;n)$ denotes the number of partitions of $n$ into parts which are not congruent to $0, \pm a\pmod b$, then for every positive integer $n$ we have $$ P_{27}(12; n)=P_{27}(6;n-1) + P_{27}(3;n-2). $$ We also show that these quotients classify supersingular elliptic curves in characteristic $p$. More precisely, if $2k+1=p$, where $p\geq 5$ is prime, and the quotient is non-zero, then it is essentially the locus of characteristic $p$ supersingular $j$-invariants in characteristic $p$.

math.QA