excluding new layer from scale_size - r

Question

I have plotted a scatter plot with the point size scaled by frequency:
g<-ggplot(d, aes(x=Treatment, y= Seam.Cell.Number, size=Frequency))+geom_point(aes(colour=Strain))+ scale_size_continuous(range = c(3, 10), breaks=c(0,1, 2, 3, 4, 5,6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50))+guides(size=FALSE)
Now I am trying to plot means with standard error bars on top. I have the mean and standard error already calculated in columns in my csv file. So so far I have attempted:
g+geom_point(aes(x=Treatment,y=Mean))+geom_errorbar(aes(ymin=Mean-Standard.Error, ymax=Mean+Standard.Error, width=.4))+theme(axis.text.x = element_blank())+theme(legend.key = element_rect(colour = "black"))
And:
g+layer(data=d, mapping=aes(x=Treatment,y=Mean), geom="point")+geom_errorbar(aes(ymin=Mean-Standard.Error, ymax=Mean+Standard.Error), width=.4)+ylab("Seam Cell Number")
But they both give me very fat error bars/data points. It seems they are being affected by my size scaling in object g. I have tried to modify the size and width of the error bars, and I have tried to modify the size of the data points, both in these last bits of code, but to no avail. Is there a way to 'cancel' the size command for this layer?

Answer 1

If you reverse the order of your ggplot, you may be able to avoid the size distortion on the error bars.
Not having reproducible data, I made some up.
df <- data.frame(Treatment = (1:100), Seam.Cell.Number = 3:102, Frequency = 5:104,
Strain = rep(c("A", "B", "C", "D"), 25))
std <- function(x) sd(x)/sqrt(length(x))
Mean <- mean(df$Treatment)
df$Standard.Error <- std(df$Treatment)
g <- ggplot(df, aes(x = Treatment, y = Seam.Cell.Number)) +
geom_point(aes(x=Treatment, y=Mean)) +
geom_errorbar(aes(ymin=Mean-df$Standard.Error, ymax=Mean+df$Standard.Error, width=.4))+
theme(axis.text.x = element_blank())+
theme(legend.key = element_rect(colour = "black"))
g + geom_point(aes(colour=Strain)) +
scale_size_continuous(range = c(3, 10), breaks=c(0,1, 2, 3, 4, 5,6, 7, 8, 9, 10, 11, 12, 13, 14, 15,
16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28,
29, 30, 31, 32, 34, 35, 36, 37, 38, 39, 40, 41, 42,
43, 44, 45, 46, 47, 48, 49, 50)) +
guides(size=FALSE)