Hello can anyone help,
I have been running the following script to obtain a PCA plot but the end
result is rather disappointing as the points are very very small and there are
no titles etc
geochemdata<-read.csv(file.choose(),header=TRUE)
names(geochemdata)
library(vegan)
bstick<-function(n, tot.var=1) rev(cumsum(tot.var/n:1)/n)
geopca<-rda(geochemdata, scale=TRUE)
geopca
geopca$CA$eig/geopca$tot.chi
summary(geopca, scaling=2)
apply(scores(geopca, choices=1.6, display= "species"),
+ 2,function(x) x/sd(x))
plot(geopca$CA$eig, type= "o", col= "red", xlab= "PCA Axis",
ylab= "Variance", main= "Scree plot for the PCA of the Geochem Data")
bstick.env<- bstick(19, tot.var=19) #must change number based on number in
bstick model
bstick.env
geopca
plot(bstick.env, type="o", lty="dotted", ylim=range(bstick.env,
geopca$CA$eig), xlab="PCA Axis", ylab="Inertia",
main="Geochemical ITRAX Data: Bstick")
points(geopca$CA$eig, type="o", col="red")
plot(geopca, scaling=2)
env.sc<- scores(geopca)$species
plot(geopca, scaling=2)
arrows(0, 0, env.sc[,1]*0.85, env.sc[,2]*0.85,
col="red", length=0.05)
Does anyone know a more suitable code? I have also tried
eochemdata<-read.csv(file.choose(),header=TRUE)
names(geochemdata)
fit <- princomp(geochemdata, cor=TRUE)
summary(fit) # print variance accounted for
loadings(fit) # pc loadings
plot(fit) # scree plot
fit$scores # the principal components
write.csv(fit$scores, file="trial.csv") # export output of PCA
which does work but again I cannot seem to adjust the output. All the titles
become squashed/un-readable and the points are in text format rather than
appear as mere dots. Is there code to allow the user to move things about as
they so wish and set their own style? I have used all the online suggestions
but they keep coming back with error comments.
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
