r error bars and y axis labels not in order-CodePudding

I am generating a plot to show the estimates and CI graphically. Right now my estimates and y axis labels are out of synch on the plot.

If this is my dataset.

       Group      Est     conf.low      conf.high   pvalue   
       Bi         1.12    0.65          1.603       0.000
       Di        -0.24   -0.44         -0.038       0.02
       Dn        -0.47   -0.80         -0.140       0.005
       HMD       -0.006  -0.32          0.311       0.968
       HMS        0.72   -1.00         -0.436       0.000
       LM        -0.055  -0.32          0.214       0.6886
       PaS       -1.31   -1.79         -0.850       0.000
       Ph A       0.065  -0.250         0.381       0.6885
       TRB        1.023   0.63          1.41        0.000
       TRC       -0.599  -0.94         -0.249       0.0008

What I am seeing is a plot like this below where the y axis labels and estimates are completely out of order.

This is my code

     ggplot(df, aes(y = row.names(df), x= Est)) 
        geom_point(shape = 16, size =5) 
          scale_y_discrete(label = df$Group)   
             geom_errobarh(aes(xmin = conf.low, xmax = conf.high), width =0, size = 1) 
               geom_vline(xintercept=1, color="black", linetype = "dashed", cex=1, alpha=0.5)

Any help on how to sync my y-axis labels with the estimates is much appreciated. Thanks.

CodePudding user response：

If you want the plot to appear in the same order as the data in your data frame, then you need to plot Group on the y axis, but first convert it to a factor whose levels are the reverse of the column Group:

library(ggplot2)

ggplot(df, aes(y = factor(Group, rev(Group)), x = Est))  
  geom_point(shape = 16, size = 5)  
  geom_text(aes(label = Est), nudge_y = 0.3, size = 4)  
  geom_errorbarh(aes(xmin = conf.low, xmax = conf.high), height = 0, size = 1)  
  geom_vline(xintercept = 1, linetype = "dashed", cex = 1, alpha = 0.5)  
  labs(y = "Group")  
  theme_gray(base_size = 16)

Reproducible data

df <- structure(list(Group = c("Bi", "Di", "Dn", "HMD", "HMS", "LM", 
"PaS", "Ph A", "TRB", "TRC"), Est = c(1.12, -0.24, -0.47, -0.006, 
-0.72, -0.055, -1.31, 0.065, 1.023, -0.599), conf.low = c(0.65, 
-0.44, -0.8, -0.32, -1, -0.32, -1.79, -0.25, 0.63, -0.94), conf.high = c(1.603, 
 -0.038, -0.14, 0.311, -0.436, 0.214, -0.85, 0.381, 1.41, -0.249
), pvalue = c(0, 0.02, 0.005, 0.968, 0, 0.6886, 0, 0.6885, 0, 
8e-04)), class = "data.frame", row.names = c(NA, -10L))

^{Created on 2022-10-29 with}

CodePudding user response：

Credits: I had forgotten the values above the points, correction copy&pasted from Allan Cameron's answer.

Here is a solution with reorder to put the y axis in order of the bars mid points, Est.

df1<-"Group      Est     conf.low      conf.high   pvalue   
       Bi         1.12    0.65          1.603       0.000
       Di        -0.24   -0.44         -0.038       0.02
       Dn        -0.47   -0.80         -0.140       0.005
       HMD       -0.006  -0.32          0.311       0.968
       HMS        -0.72   -1.00         -0.436       0.000
       LM        -0.055  -0.32          0.214       0.6886
       PaS       -1.31   -1.79         -0.850       0.000
       'Ph A'       0.065  -0.250         0.381       0.6885
       TRB        1.023   0.63          1.41        0.000
       TRC       -0.599  -0.94         -0.249       0.0008"
df1 <- read.table(textConnection(df1), header = TRUE)

library(ggplot2)

ggplot(df1, aes(x = Est, y = reorder(Group, -Est)))  
  geom_pointrange(aes(xmin = conf.low, xmax = conf.high), size = 1)  
  geom_text(aes(label = Est), nudge_y = 0.3, size = 4)  
  geom_vline(xintercept = 1, linetype = "dashed", alpha = 0.5)  
  ylab("Group")

^{Created on 2022-10-29 with reprex v2.0.2}

And with error bars, instead of line bars. This time the points need an explicit call to geom_point.

ggplot(df1, aes(x = Est, y = reorder(Group, -Est)))  
  geom_point(pch = 16, size = 5)  
  geom_errorbarh(aes(xmin = conf.low, xmax = conf.high), height = 0.5, size = 1)  
  geom_text(aes(label = Est), nudge_y = 0.3, size = 4)  
  geom_vline(xintercept = 1, linetype = "dashed", alpha = 0.5)  
  ylab("Group")

^{Created on 2022-10-29 with reprex v2.0.2}