- #1
FallenApple
- 566
- 61
Would the following method work? I could uniformly distribute points into the data cloud. Of the "darts" that I threw in, create edges between all points with a metric value under a certain amount. The nodes in the resulting graph that have more neighbors would indicate greater density. I could tune the amount of darts and since the connective density would increase in a more non linear fashion for denser regions vs less dense regions, I could tell if there's a cavity or not.
Would this work in practice?
Would it work topologically? That is, I would get the same indication for a coffee mug cloud vs a doughnut? I mean, regardless of the shape, I would end up with two less dense regions.
Would this work in practice?
Would it work topologically? That is, I would get the same indication for a coffee mug cloud vs a doughnut? I mean, regardless of the shape, I would end up with two less dense regions.