Figure 1
Examples of circular point clouds with radius r = 0.5 generated with one true-random and two quasi-random (Sobol, Halton) filling algorithms. Each cloud is generated by initially filling 5000 points into a squared area with side length l = 1 and subsequently deleting all points outside of the circle, which leaves ∼4000 points. The usage of a quasi-random algorithm leads to a higher homogeneity of the spatial distribution relative to the true-random method, whereas the true-random distribution shows a higher degree of local clustering. |