Hi, 

I would like to use my data on distances between consecutive captures of a butterfly for constructing negative exponential function and inverse power function that would estimate the probability for long distance movements. 

Although the procedure seems quite straight-forward, I am strugelling with it. It is often told that one should use inverse cumulative proportions of individuals moving a ceartain distance class (e.g. 100 m) and to make their regression on distances in NEF (negative exponential function) or natural logarithms of distances in IPF (inverse power function). The problem that occurs is that the cumulative proportion for the larger distance class is one, and using logarithm on it always result in zero. I suspect that for that reason, the statistical program (I am using SPSS) rejects the curve estimation. I may be wrong with the latter conclusion (or I am doing something wrong in any of the early steps) but I would be very grateful for any useful suggestions, and maybe some references on that topic.

Kind regards,

Jure