Caterpilar Plot in R

Question

I'm trying to make a Caterpillar plot from the confident intervals of a series of glmm outputs. So, I made a simple table (tabla) with the Index names (8) the Parameters (3 for each index) and the Intervals (2 values for each Parameter = min_value - max_value). I 'd like my plot to be grouped by Index and by the 3 parameters of each index in the Y-axis.

I tried out with ggplot but my poor knowledge did not let me continue.

my code :

library(ggplot2)
read.csv("ggplot_modelos.csv", header = TRUE, sep = ";", dec = ".")-> tabla

View(tabla)

names(tabla)

ggplot(tabla, aes(Intervalls, Index) , geom_bar(aes(Intervals, Parameter)))

My data:

tabla <-
data.frame(
  stringsAsFactors = FALSE,
             Index = c("H","H","H","H","H","H",
                       "ADI","ADI","ADI","ADI","ADI","ADI","AEI","AEI",
                       "AEI","AEI","AEI","AEI","BIO","BIO","BIO","BIO",
                       "BIO","BIO","ACI","ACI","ACI","ACI","ACI","ACI","M",
                       "M","M","M","M","M","Ht","Ht","Ht","Ht","Ht",
                       "Ht","AR","AR","AR","AR","AR","AR"),
         Parameter = c("Intercept","Intercept",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","Intercept","Intercept",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","(Intercept)","(Intercept)",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","(Intercept)","(Intercept)",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","(Intercept)","(Intercept)",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","(Intercept)","(Intercept)",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","(Intercept)","(Intercept)",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance","(Intercept)","(Intercept)",
                       "Species Richness","Species Richness","Relative Abundance",
                       "Relative Abundance"),
        Intervalls = c(-0.63399094,0.10088021,
                       0.05111166,0.11374545,0.01356644,0.03568559,-0.63399094,
                       0.10088021,0.05111166,0.11374545,0.01356644,0.03568559,
                       -0.08921982,0.69843374,-0.11735036,-0.05706148,
                       -0.03492262,-0.01363224,-0.387777431,0.8894038,-0.01656506,
                       0.04067182,-0.004208021,0.01583599,-1.42348399,
                       0.78640078,0.13746715,0.21376064,0.06880861,0.09518447,
                       -0.44656228,0.46793251,0.03039887,0.05458163,
                       0.05991785,0.12952663,-0.384061,1.50738256,-0.2298068,
                       -0.1467239,-0.106336,-0.07786533,-0.54823458,1.23599168,
                       -0.04350877,-0.01813159,-0.09811032,-0.02568451)
)

I will appreciate if anyone can give me a hand here!

Typically it's `ggplot(data, aes(x, y))` first (notice the parens close), and then add (`+`) each component/layer/theme element, such as your `geom_bar`. Over-simplified, it often looks like `ggplot(...,aes(...)) + geom_bar(...)`. — r2evans, Jul 29 '20 at 22:35
Maybe this can help: https://stackoverflow.com/questions/34344599/a-caterpillar-plot-of-just-the-significant-random-effects-from-a-mixed-effects/34345877#34345877. And are you sure you want `geom_bar`? — Edward, Jul 29 '20 at 22:40

score 0 · Accepted Answer · answered Jul 30 '20 at 00:22

Your data needs reshaped before plotting. The lower and upper end of the ranges should be in different columns rather than consecutive rows. You probably also want to use segments instead of columns:

library(tidyr)
library(dplyr)
library(ggplot2)

p <- tabla %>%
  mutate(type = rep(c("min", "max"), nrow(tabla)/2)) %>%
  pivot_wider(names_from = type, values_from = Intervalls) %>%
  mutate(Parameter = gsub("\\(Intercept\\)", "Intercept", Parameter)) %>%
  ggplot() + 
  geom_segment(aes(y = Parameter, yend = Parameter, x = min, xend = max,
                   colour = Parameter), size = 2) +
  scale_color_manual(values = c("gray90", "deepskyblue4", "orange")) +
  facet_grid(Index~., switch = "y") +
  scale_y_discrete(expand = c(0.5, 0.5)) +
  theme_classic() +
  labs(x = "Coefficient") +
  theme(panel.grid.major.x = element_line(),
        strip.text.y.left = element_text(angle = 0, size = 14, face = 2),
        strip.placement = "outside",
        strip.background = element_blank(),
        axis.text.y = element_blank())

p

You might also want to "zoom in" on the x axis to see the coefficients better while sacrificing being able to see the full intercept ranges:

p + coord_cartesian(xlim = c(-0.25, 0.25))

great Allan!! thank you very much. you clarified my plot – Martin Jul 30 '20 at 13:07 — Martin, Jul 30 '20 at 13:07

Caterpilar Plot in R

1 Answers1