plot group means using multiple columns in ggplot2 (R)

Question

I have a dataset that has a group column and a parameter estimate column for that group. I want to plot the estimates and CIs on the yaxis, and have separate rows for each group and parameter estimate. I know that I can create a composite column from the group and paramter esimate column, but I am wondering if there is a way in ggplot2 to keep the columns separate.

Here is some example code.

group1 = c(rep("A",3),rep("B",3),rep("C",3))
group2 = c(rep(c("param1","param2","param3"),3))
est = rnorm(9,mean = 0, sd = 1)
lwr = est - sd(est)
upr = est + sd(est)

df = data.frame(group1,group2,est,lwr,upr)

figure.gg = ggplot(data = df, aes(x = group2, y = est, ymin = lwr, ymax = upr)) +
  geom_point(position = position_dodge(width = 0.5)) +
  geom_errorbar(position = position_dodge(width = 0.5), width = 0.1) +
  coord_flip() +
  ylab("estimate")

figure.gg

Joel Kandiah · Accepted Answer · 2021-04-05 15:23:35Z

I hope I have understood this correctly please add a comment to clarify if this is not the case.

One solution would be to use facets to separate the plot the rows. To do this I have used the facet_wrap() function to separate the different group 2s and set the y axis to use the different group 1s.

library(tidyverse)

group1 = c(rep("A",3),rep("B",3),rep("C",3))
group2 = c(rep(c("param1","param2","param3"),3))
est = rnorm(9,mean = 0, sd = 1)
lwr = est - sd(est)
upr = est + sd(est)

df = data.frame(group1,group2,est,lwr,upr)

# Swapped the x axis to use group1
figure.gg = ggplot(data = df, aes(x = group1, y = est, ymin = lwr, ymax = upr)) +
  geom_point(position = position_dodge(width = 0.5)) +
  geom_errorbar(position = position_dodge(width = 0.5), width = 0.1) +
  coord_flip() +
  # Facet wrapped with one column using group 2s
  facet_wrap(~group2, ncol = 1, strip.position = "right") +
  ylab("estimate")

figure.gg

^{Created on 2021-04-05 by the reprex package (v2.0.0)}

Alternatively we could use a secondary grouping to separate the different entries such as group = group1 or colour = group1 which are set within aes(). These are presented below


# Set group 1 as a group in aes
figure.gg = ggplot(data = df, aes(x = group2, group = group1, y = est, ymin = lwr, ymax = upr)) +
  geom_point(position = position_dodge(width = 0.5)) +
  geom_errorbar(position = position_dodge(width = 0.5), width = 0.1) +
  coord_flip() +
  ylab("estimate")

figure.gg


# Set group1 as the colour using aes
figure.gg = ggplot(data = df, aes(x = group2, colour = group1, y = est, ymin = lwr, ymax = upr)) +
  geom_point(position = position_dodge(width = 0.5)) +
  geom_errorbar(position = position_dodge(width = 0.5), width = 0.1) +
  coord_flip() +
  ylab("estimate")

figure.gg

^{Created on 2021-04-05 by the reprex package (v2.0.0)}

Thanks. The first bit of code you have using facet_wrap was exactly what I was looking for. My only problem now is that I actually have some unused levels and need to drop them. Not sure how to do that with facet_wrap, as drop = TRUE is not working. I will ask as a separate question.

TarJae · Accepted Answer · 2021-04-05 15:46:48Z

0

I think this makes more sense:

ggplot(df1, aes(x=group1, y=est, colour=group2)) + 
  geom_errorbar(aes(ymin=lwr, ymax=upr), position = position_dodge(width = 0.5), width = 0.1) +
  geom_point(position = position_dodge(width = 0.5)) +
  theme_bw()

answered Apr 5, 2021 at 15:46

TarJae

80.2k6 gold badges30 silver badges94 bronze badges

Collectives™ on Stack Overflow

plot group means using multiple columns in ggplot2 (R)

2 Answers 2

1 Comment

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related