I'm interested in performing Principal Coordinate Analysis (PCoA) to plot the functional trait space of plants based on e.g. fruit colour, fruit size, plant height, and compare one group of plants (e.g. one island) to another group.
I've been having difficulties finding example R code to perform PCoA, and was wondering if anyone could provide an example with R code on how to run a PCoA?
Thank you!
Hi Wilfredo!
You need two sets of data, community data matrix, and environmental data matrix. Sample data from the environmental data (env) and community data matrix should be arranged in the same order, site 1 , site 2, etc.
I presume you have transformed your data, and calculated the dissimilarity distance of as appropriate from the community data matrix.
I use the vegan package, and here is an example, which I am sure can be improved.
library(vegan)
pco.result
Hola Wilfrido,
I give you a code which is based in the Book "Numerical Ecology with R" (Borcard et al 2011). I recomend you to read the book to get into details. Note:
"spe" is your matrix of species (columns species, rows sites where they were found)
# *********************************************************************
# Load the required packages
# (vegan must be loaded after ade4 to avoid some conflicts)
library(ade4)
library(vegan)
library(gclus)
library(ape)
library(FactoMineR)
# PCoA on a percentage difference (Bray-Curtis) dissimilarity matrix of species
spe.bray
Hi, Feel free to give a look to the web page développer by Thiago G. Martins at https://tgmstat.wordpress.com/2013/11/28/computing-and-visualizing-pca-in-r/
He used an exemple of data (iris)
Presented how to transform the data (log-transformation)
computed the function pca, with the scale
print the result to comment and see the evolution of the different parameters measured.
finally very nice plot computed with ggplot
@ Jorge Antonio Gomez-Diaz I hope you´re doing great.
Can you show me an example script to make the plot in a most publishable format?
Specifically, for these step of the script you shared:
# First biplot: Hellinger-transformed species data
biplot.pcoa(all.pa.h.pcoa, all.pa.h, dir.axis1=-1)
abline(h=0, lty=3)
abline(v=0, lty=3)
# Second biplot: standardized Hellinger-transformed species data
all.pa.std
how to solve this error
Error in wascores(spe.b.pcoa$points[, 1:2], data) :
weights must be non-negative and not all zero
- Badges
- Science topic
- Similar topics
- Filing