This software is implemented as part of the course Computational Geometry (Fall 2014) at Aarhus Univerity Denmark.
Author: Martin Storgaard and Konstantinos Mampentzidis
To create an executable for running either convex hull algorithm use: make main Afterwards execute ./main [alg] [inputFile] [outputFile], where [alg] is a number between 1 and 4: 1: Incremental CH 2: Divide-and-Conquer CH 3: MbC without extra pruning 4: MbC with extra pruning and [inputFile] and [outputFile] are paths to the input file and output file respectively.
To generate the test cases used in our report, simply use make tests (Warning: This might take approximatly 20 hours).
When all tests cases are created, one can run performance benchmark using make perf, and then execute ./perf
To compare the output with qhull, simply runn ./correctness.py - this script compares all input data located in data/
To generate other shapes, see python generators/[shape].py -h, where shape is either, circle curve, lines, square or triangle.