Hard in Theory, Easy in Practice: Solving Large Minimum-Weight Triangulation Instances to Provable Optimality - Non-random Instance Dataset

The non-random point sets used in our work on Minimum Weight Triangulation. Among instances generated to have multiple components in their LMT skeleton and other constructed instances, contains instances from various sources (but potentially brought into a different format), including the points from TSPLIB instances (http://comopt.ifi.uni-heidelberg.de/software/TSPLIB95/), VLSI instances (https://www.math.uwaterloo.ca/tsp/vlsi/index.html) and some point sets of Salzburg polygon database polygons (https://zenodo.org/records/3784789).本研究用于最小权三角剖分（Minimum Weight Triangulation）的非随机点集数据集。本数据集涵盖三类实例：其一为生成得到的、其LMT骨架（LMT skeleton）包含多个连通分量的实例；其二为其他人工构造的实例；其三则来自多种不同来源（但格式可能经过统一调整），包括旅行商问题库（TSPLIB）实例（http://comopt.ifi.uni-heidelberg.de/software/TSPLIB95/）中的点集、超大规模集成电路（VLSI）实例（https://www.math.uwaterloo.ca/tsp/vlsi/index.html）中的点集，以及萨尔茨堡多边形数据库（Salzburg polygon database）中部分多边形对应的点集（https://zenodo.org/records/3784789）。