Sample based random swapping algo is identical to individual rarefaction about mobr HOT 3 CLOSED

The above figure was generated using a simple simulated community in which I attempted to only change N (some small SAD effects crept in as well). On the left panel there are four types of curves:

1. individual based rarefaction curve (note they are cut off a minimum number of individuals so the high density curve is truncated)
2. a sample based rarefaction curve in which the abundance of each species is redistributed randomly across plots (i.e., this is in essence again an individual based rarefaction curve)
3. the rescaled sample based curve using average density of the two groups.
4. the interpolated rescaled sampled based curve for 1 to 9 individuals so that delta S can be computed.

The blue curves are for the community with higher density, the red curves are for the community with lower density. The delta S plotted in the right panel is the effect of N within each group we just need to compute the differences between these two curves to estimate the effect of N on differences between treatment richness.

from mobr.

These changes have been incorporated into #60

from mobr.

Hey @dmcglinn - thanks for all the hard work! I'm back in US now and slowly catching up.

Your insight makes perfect sense, I'm glad you figured it out (no more swapping needed, yay!). So I guess the effect of N comes purely from our operation of swapping Nplots with Nind (which also makes sense)?

I'll try to review your pull request tonight.

from mobr.