I want to do a PCA (in R) but I dont know if my data is linear. I conducted an NMDS for Eucladian and Bray-Curtis distance. After removal of 1 variable with many zeros, the NMDS plots are roughly equal.
My assumption is that linear data plotted in non-linear ordination space (Bray-Curtis) results in similar results as linear data plotted in linear ordination space (Euclidian). Is this assumption justified?
Thanks!!