Maybe this question is not in the right place and if so, I'll delete it.
Probably a very basic question: If NMDS (vegan package) convergence does not reach a solution (regardless of dimensions and iterations) does that make the output meaningless?
Because R does give an output. With a nice low stress level, good plot. All of that.
Thank you.
You could try to change the number of dimensions (k = 3 or more), but display only the two first dimensions.
Some dataset are too complex to reach a solution in a 2d space. Even after trymax=100000 one of my 'sp x site' dataset did not reached a solution. However, the stress kept consistently low in all iterations, then I assumed that no best solution was possible. Maybe after billions of iterations.
