Direct integration for general \(\Omega\) backgrounds (Q1951043): Difference between revisions

\(Z(\epsilon_1,\epsilon_2)\), the instanton partition function of an \(N=2\) gauge theory in the general \(\Omega\) background, resp., the refined BPS partition function of a local Calabi-Yau threefold, is studied from the B-model point of view. The coefficient functions \(F^{(\frac n2,g)}\) in the 2-parameter asymptotic expansion, \[ \log Z(\epsilon_1,\epsilon_2) \sim \sum_{g,n} (\epsilon_1+\epsilon_2)^n (\epsilon_1 \epsilon_2)^{g-1} F^{(\frac n2,g)} \] satisfy a ``generalized'' holomorphic anomaly equation over the moduli space of vacua (whose rigid special geometry is encoded in \(F^{(0,0)}\)). That equation, being equivalent to the ordinary holomorphic anomaly equation (see [\textit{D.~Krefl} and \textit{J.~Walcher}, Lett. Math. Phys. 95, No. 1, 67--88 (2011; Zbl 1205.81118)]), can be solved by the same methods developed by the authors in previous works, such as, direct integration and the (generalized) gap condition at the monopole/dyon/conifold points. The formalism is applied to a variety of interesting \(N=2\) theories, such as the pure \(SU(2)\) Seiberg-Witten theory, and ``refined topological strings'' on local half K3, local \({\mathbb P}^1\times {\mathbb P}^1\), and local \({\mathbb P}^2\).