In ggplot2, generate error bars across Facets in a data with multiple independent variables

Question

I am trying to put error bar on mean values from a data frame which has three independent variables and plotted as facet_grid. However, the plot below is putting error bars in wrong facets. Could anyone please help me?

Please see below the example data and associated code:

life <- rep(c("1d", "2d", "4d"), 2, each = 2)
trt <- rep(c("c1", "c2"), 6)
species <- rep(c("SP1", "SP2"), each = 6)
mean_v <- runif(12, 12, 45)
sem_v <- runif(12, 1, 4)
data1 <- data.frame(life, trt, species, mean_v, sem_v)
plot1 <- ggplot(data1, aes(x = trt, y = mean_v, group = species, fill = species))
plot1 + geom_bar(stat = "identity", position = "dodge") + 
        facet_grid(~life) + 
        geom_errorbar(aes(ymin = data1$mean_v - data1$sem_v, 
                          ymax = data1$mean_v + data1$sem_v, 
                          width = 0.2),
                      position = position_dodge(width = 0.90), 
                      group = data1$trt)

Thanks very much in advance.

Answer 1

The solution seems to be to specify position=position_dodge(width=0.9) in both geom_bar and geom_errorbar.

library(ggplot2)

plot1 <- ggplot(data1, aes(x=trt, y=mean_v, group=species, fill=species)) +
         geom_bar(stat="identity", position=position_dodge(width=0.9)) + 
         facet_grid(. ~ life) + 
         geom_errorbar(aes(ymin=mean_v - sem_v, ymax=mean_v + sem_v), 
                       width=0.2, position=position_dodge(width=0.9))

ggsave("dodged_barplot.png", plot=plot1, height=4, width=6, dpi=150)