ggplot2 produces error when used in function or S4

Question

I\'ve tough time with ggplot2 when used in function or S4. Here is my code without function:

rm(list=ls(all=TRUE))
library(nlme)
data(Oats)
Data <- Oats

         


        

Answer 1

Just add a label columns to the data frames. Also, note that you should never have a $ inside aes.

This should work:

G1 <- data.frame(G, label=rownames(G))
E1 <- data.frame(E, label=rownames(E))

library(ggplot2)
r <- 0.08
p <- ggplot(data=G1, aes(PC1, PC2)) + geom_point() + theme_bw()
p <- p + geom_text(aes(label=label), size=3, vjust=1.25, colour="black")
p <- p + geom_path(data=as.data.frame(con.hull), aes(PC1, PC2))
p <- p + geom_segment(data=as.data.frame(E), aes(xend = PC1, yend=PC2),
                      x=0, y=0, colour="black",
                      arrow=arrow(angle=25, length=unit(0.25, "cm")))
p <- p + geom_text(data=E1, aes(label=label),
                   size=3, vjust=1.35, colour="black")
p <- p + labs(list(x=sprintf("PC1 (%.1f%%)", PC.Percent.SS[1]),
                   y=sprintf("PC2 (%.1f%%)", PC.Percent.SS[2])))
p <- p  + opts(axis.title.x=theme_text(size=10, hjust=0.54, vjust=0))
p <- p  + opts(axis.title.y=theme_text(size=10, angle=90, vjust=0.25))
p <- p + xlim(range(c(E[ ,1], G[ ,1])) + range(c(E[ ,1], G[ ,1]))*r)
p <- p + ylim(range(c(E[ ,2], G[ ,2])) + range(c(E[ ,2], G[ ,2]))*r)
p <- p + geom_segment(data=tmp, aes(xend=V1, yend=V2), x=0, y=0)
print(p)

Answer 2

You have to set as.data.frame(G) outside of your ggplot2 call if you want to be able to use rownames(G).

Set data=as.data.frame(G) in your call to geom_text.

Answer 3

p <- p + geom_text(aes(label = row.names(G)), size = 3, vjust = 1.25, colour = "black")

You're providing it with a layer, but no data to plot. Either move label= into your original call to ggplot or provide geom_text() with a data= argument

Answer 4

Your version in the function is missing the following two lines at the very beginning compared to the original version:

Data$Env <- factor(Data$Block)
Data$Gen  <- factor(Data$Variety)

Have you experimented with using debug() to step through functions to find the source of errors?