On Friday, April 5, 2013 8:22:05 AM UTC+13, K wrote: > Another possibility: It seems that if I numerically integrate the two types of data using the trapz() function, I get distinct results. Upward trending data have lower integrals than downward trending data. The Y axis is always fixed between 1 and 100 for all points, so this may work well. > > > > K
Have you considered that a reason no one is responding to your posts is that people (myself included) do not understand what you're getting at? That's why I asked for the original data. Not because I wanted to know where they came from, but because I wanted to know what they looked like and what you meant by upward and downward trends. In my experience, detecting trends is best done by differentiating, not integrating, but what do I know?