add autocorrelation estimate of effective number of trials to DQReport
We should estimate the autocorrelation length of the timeseries and then leverage this into an "effective number of trials" within some window. From that, we can estimate the overall probablity of seeing something with FAP <= what was observed somewhere within the window as a distribution of the extremum of a uniformly distributed variable.
That should be what we want to answer the question "how likely is it that iDQ does this thing within some random window," but perhaps not. Since p(glitch|aux) is (close to) 1-to-1 with FAP (by virtue of the calibration map), we may be able to estimate the probablity of seeing a deviation in FAP and call that the probablity of seeing a deviation in p(glitch|aux), but again I'm not sure. We need to think about this more carefully.