My problem is I have tens of thousands of points (GPS) representing animal locations over a period of several months. I want to test if a portion of those points (only about a few hundred) which occurred during stressful events have a markedly different distribution than the overall distribution of the remainder points.
What method/statistical route should I use? I've heard of Ripley's K Function, but I don't know enough about it to know if it's applicable, especially with the relatively small amount of points representing the disturbed case. The trouble is also that many of these methods sound like they indicate if a distribution is clustered or random but not whether or not distributions seem to match.
For this test I would have time involved but I will most likely assume that their behavior is not seasonality-dependent (although I know inevitably it is)
If it's possible I would also like to compare whether the directions the animals tended to move varied between the two circumstances. So normally the animals follow a particular pathway and locations off of the pathway tended to result in motion towards the pathway, I want to know if I can test to see if the disturbed points tended to have their directionality (as determined by their next subsequent point) no longer fit that pattern, i.e. avoiding the pathway.