I'm working with a microbiome dataset with 3246 variables and 107 observations. I'm using metaMDS to do an NMDS plot with the default parameters.
Even though I reached a solution there are no site scores.
> ord <- metaMDS(data_dist)
Run 0 stress 0.1888251
Run 1 stress 0.1911371
Run 2 stress 0.188518
... New best solution
... Procrustes: rmse 0.03850648 max resid 0.2697121
Run 3 stress 0.196947
Run 4 stress 0.2013088
Run 5 stress 0.2017199
Run 6 stress 0.1911485
Run 7 stress 0.1888848
... Procrustes: rmse 0.03362846 max resid 0.2705997
Run 8 stress 0.2034295
Run 9 stress 0.1890825
Run 10 stress 0.1928713
Run 11 stress 0.1907301
Run 12 stress 0.1904192
Run 13 stress 0.1898112
Run 14 stress 0.1962496
Run 15 stress 0.1898825
Run 16 stress 0.1927416
Run 17 stress 0.1885187
... Procrustes: rmse 0.0001846372 max resid 0.001710078
... Similar to previous best
Run 18 stress 0.1947141
Run 19 stress 0.1995386
Run 20 stress 0.1904549
*** Solution reached
> score <- scores(ord, display = "sites")
Error in class(S) <- "scores" : attempt to set an attribute on NULL
I tried removing variables for which there are only one positive observation.
checkNumZerosCol <- apply(data_dist,2,function(x) sum(x==0))
cases <- which(checkNumZerosCol == (nrow(data_dist) - 1))
length(cases)
data_dist <- data_dist[,-cases]
I tried scaling the data to have a range between 0 and 10 (which I think metaMDS already does by itself...)
data_dist[] <- data_dist^0.25
I tried calculating the dissimilarity matrix before the ordination
dist <- vegdist(data_dist, method = "bray")
I used metaMDS so many times and it always worked before, I really don't know what to do to get those site scores... Does anyone know what's wrong with this community matrix?
Here is my community matrix : https://drive.google.com/file/d/1awS9xew241JsPABUc4NH9sY-YW2-_a23/view?usp=sharing
