rqResMar_dist: Marginal Randomized Quantile Residuals
In philipshirk/nmmsims: Simulate & Analyze N-mixture Model Data

Calculates Marginal Randomized-quantile residuals for a distance model fit with unmarked. Note that I currently calculate these residuals by subtracting 1 from each distance bin. That's pretty hacky and likely just plain wrong. Don't use residuals from more than 1 distance bin. Just use residuals from a single bin. It would be better to extend the observation rq residuals to a multinomial distribution, and use the pbinsum function to get the possible N's (up to K) across all distance breaks. But then I could just draw a random distance from the normal distribution, take 1 individual out of that distance category (for the F(z-1) calculation) and calculate the residuals with those 2 values & distributions. Use rqResMar function for point-count model. As defined in Knape et al. 2018 and available for unmarked models in the nmixgof package.