Does the metaMDS function in vegan rotate the ordination solution so the first axis explains the most variance? If not, is there a way to achieve this?
Run 20 stress 0.09957583
... Procrustes: rmse 0.0001349268 max resid 0.0009665635
... Similar to previous best
I am also unsure about how to interpret the procrustes data. What do the values for RMSE and max residual represent?
Thanks!
Yes, it does. Documentation (?metaMDS) expresses it this way:
Principal components rotate the configuration so that the variance of points is maximized on first dimension
The output also prints when PC rotation was used:
Scaling: centring, PC rotation, halfchange scaling
However, this has nothing to do with Procrustes rotation that is used to asses the similarity of solutions during iteration steps. PC rotation only concerns the final returned result.
About interpretation of the RMSE and max residual: they are statistics that compare two solutions during iteration. RMSE is a kind of average difference, and max is the maximum difference. If you they are small, two iterations yielded similar results. To see what they mean, run metaMDS with option plot = TRUE and all Procrustes rotation are plotted when running the results. The blue arrows in that plot will show the differences, and RMSE is an averaged arrow length, and max residual is the length of the longest arrow. If you won't see much arrows, then two solutions are so similar that differences are not visible.
