Package dev.nm.solver.multivariate.constrained.convex.sdp.socp.interiorpoint
package dev.nm.solver.multivariate.constrained.convex.sdp.socp.interiorpoint
-
ClassesClassDescriptionThis implementation is based on Algorithm 14.5 in the reference.Solves a Dual Second Order Conic Programming problem using the Primal Dual Interior Point algorithm.The SOCP dual problem we are solving here is : \max {\bm b}^T \hat{\bm y} \\ {\rm s.t.} ({\bm A_i^q})^T \hat{\bm y} + {\bm z_i^q} = c_i^q,\ {\bm z_i^q}\in \mathcal{K}_q^{q_i},\ for i\in [n_q];\\ ({\bm A^{\ell}})^T \hat{\bm y} + {\bm z}^{\ell} = c^{\ell},\ {\bm z}^{\ell} \ge 0;\\ ({\bm A^u})^T \hat{\bm y} = c^u;\\ \hat{\bm y} \in \mathbb{R}^m;\ {\bm z}^{\ell}\in \mathbb{R}^{n_{\ell}};\ {\bm z}^u \in \mathbb{R}^{n_u}.The vector set {x, s, y} is a solution to both the primal and dual SOCP problems.This implements Algorithm_IPC, the SOCP interior point algorithm in SDPT3 version 4.This implements Algorithm_IPC, the SOCP interior point algorithm in SDPT3 version 4.This implements Algorithm_IPC, the SOCP interior point algorithm in SDPT3 version 4.