使用 ggplot 将误差线添加到 R 中的聚集条形图

Question

I am trying to create a clustered bar graph with error bars for my independent variables (y axis) for each of my dependent variables (metrics on the x axis).

My data:

df <- data.frame (Parameter_Estimate=c('Burnham','Calumet','Northerly','Orland','Hickory Hammock','Lake O','MacArthur','Site E','Corrales','Oxbow','Tingley','Galilee','Jacobs Point','Prudence Island','Town Pond','Trial 1','Trial 2','2017','2018','Spring','Summer','0.5m','1m','Chicago','Florida','New Mexico','Rhode Island'),
     Species.Richness=c(79.4, -20.6, -12.6, -4, -63.15, -66.4, -69.15, -70.65, -52.07, -36.07, -67.23, -67.98, -69.9, -70.53, -74.73, 30.877, -2.743, 29.346, 0.9053, 29.721, -0.266, 28.898, 2.697, 70.1, -58.04, -42.49, -61.51),
     FQI=c(29.272,-0.271,-1.095,-2.652,-20.24,-19.66,-21.2,-22.56,-10.36,-5.158,-18.66,-8.802,-8.431,-7.372,-15.25,18.474,-0.14,17.869,2.1378,17.809,0.9189,18.155,1.004,28.268,-19.91,-10.39,-9.196),
     MeanC=c(3.294,0.4964,0.2252,-0.212,-0.631,-0.299,-0.339,-0.758,0.3185,0.3831,-0.135,2.7739,3.5218,4.089,3.1736,4.0308,0.1061,4.003,0.3103,4.0638,0.0266,4.0943,-0.052,3.4215,-0.634,0.0615,3.1985),
     Per.Non.Nat=c(11.912,-3.756,7.666,10.192,14.888,71.874,-4.33,-3.784,-6.288,9.517,3.31,-10.91,-11.65,-11.88,-6.388,16.086,-1.423,14.33,4.372,12.659,4.25,15.582,-0.702,15.437,16.136,-1.346,13.58),
     Shannon= c(3.6465,-0.449,-0.216,-0.4,-2.219,-2.897,-2.712,-2.504,-1.507,-0.798,-2.428,-2.066,-2.233,-2.528,-3.289,1.9313,-0.052,1.8939,0.268,1.8433,0.0975,1.8852,0.086,3.3804,-2.317,-1.311,-2.263),
     Weight.MeanC=c(3.3643,1.0304,0.4379,-0.428,-1.906,-2.679,-1.352,-1.106,1.8142,0.7625,0.1685,4.6151,5.1463,4.3978,3.6472,4.3178,0.0385,4.4046,-0.273,4.3543,-0.028,4.3357,-0.011,3.6243,-2.021,0.6551,4.1965),
     SE1=c(2.697,3.814,3.814,3.814,4.045,4.045,4.045,4.045,3.21,3.21,3.21,3.21,3.21,3.438,3.21,12.194,1.89,12.185,2.2041,12.299,2.0314,12.394,2.271,2.249,3.374,2.805,2.713))

Where SE1 is the standard error for each of my independent variables' data points for Species Richness (I haven't had the chance to add the other SE values for my other metrics, I'm just trying to experiment with the one for now).

I've created a figure using this code:

mm <- melt (df, id.vars='Parameter_Estimate')

ggplot (mm, aes (x=Parameter_Estimate, y=value) + geom_bar (stat='identity') + coord_flip()+facet_grid (.~variable) + labs (x=',y=')+scale_x_discrete (limits=df$Parameter_Estimate) +
  geom_errorbarh (data=df, aes (y=Species.Richness, xmin=SE1, xmax=SE1), size=5, color="blue",  inherit.aes = FALSE)

The barplot is not correct (SE1 is now one of the metrics along the x axis), and the error bars are not following my data for Species Richness.

Figure:

I'm unsure of what I'm doing wrong with the error bars and how to fix it!

Answer 1

You can remove the extra SE1 column by excluding it from melt:

mm <- melt (df[,-ncol(df)], id.vars='Parameter_Estimate')

Then for the error bar to appear in the Species Richness facet, you need to specify variable for it, with same levels as mm above:

se_df = df[,c("Parameter_Estimate","Species.Richness","SE1")]
se_df$variable="Species.Richness"
se_df$variable=factor(se_df$variable,levels=levels(mm$variable))

Now we plot:

ggplot (mm, aes (x=Parameter_Estimate, y=value)) + 
geom_bar (stat='identity') + 
facet_grid (~variable) + labs (x='',y='')+
geom_errorbar(data=se_df, inherit.aes=FALSE,aes (x=Parameter_Estimate,
ymin=Species.Richness-SE1, ymax=Species.Richness+SE1),
size=1, color="blue")+
scale_x_discrete(limits=df$Parameter_Estimate)+
coord_flip()