ggplot boxplot with data points: how to control both point COLOR and SHAPE based on two factors?

Question

I can successfully assign color to one factor (sex) levels. But I cannot figure out how to also assign shape to other factor (age_cat) levels and add a second legend at the desired location. I tried adding 'shape=age_cat' at various locations... But whatever I tried, distorts the plot!

My sample data, code and plot are below.

df = data.frame(sex = c(1,1,1,1,1, 2,2,2,2,2),
                age_cat = c(1,1,1, 2,2,2,  1,1,1, 2),
                score_type = c(1,2, 1,2, 1,2, 1,2, 1,2),
                score = c(25,28,18,20,30, 37,40,35,43,45))

df$sex <- factor((df$sex))
df$age_cat <- factor((df$age_cat))
df$score_type <- factor((df$score_type))

windows(width=7, height=7 )

df %>%
  ggplot( aes(x=score_type, y=score)) +
  geom_boxplot(aes(color=sex),outlier.shape = NA, size=1.5, show.legend = F) +
  geom_point(aes(color=sex), position=position_jitterdodge(dodge.width=0.9), size=3) +
  scale_color_manual(values=c("#0072B2", "#CC79A7"), name="", labels=c("Male", "Female")) +
  
  scale_shape_manual(name="", labels=c('Younger', 'Older'), values=c(1, 17)) +   # ??? How to assign shape ???
  
  theme_bw()+
  theme(panel.border = element_blank(), axis.ticks = element_blank(),
        legend.position=c(0.9, 0.65), legend.text=element_text(size=11), legend.title=element_text(size=11.5),
        panel.grid.major.x = element_blank() ,
        plot.title = element_text(size=11, face = "bold"), axis.title=element_text(size=13),
        axis.text.y = element_text(size=11),
        axis.text.x = element_text(size=11),
        plot.margin = unit(c(0.5,0.2,0,0.2), "cm")) +
  
  labs(title= "", x = "",y = "Score") +
  scale_y_continuous(breaks=c(0, 20, 40, 60, 80, 100), labels=c('0', '20', '40', '60', '80', '100')) +
  expand_limits(x=5, y=70) +
  scale_x_discrete(labels = c("A", "B")) + 
 
  coord_cartesian(clip = "off")

Maybe you clarify what you mean by "... distorts the plot!"? If you want to have different shape without affecting how the points are dodged you could do geom_point(aes(color = sex, shape = age_cat, group = sex). — stefan
– stefan, Commented Jan 12, 2022 at 23:10
This helps! Data points from factor sex remain within the boundaries of the boxplots where they should belong (the plot is not distorted). — user7395965
– user7395965, Commented Jan 12, 2022 at 23:54

Dion Groothof · Accepted Answer · 2022-01-12 23:59:11Z

1

You only need to specify shape = age_cat and group = sex in the aestetics of geom_point.

library(purrr)
library(ggplot2)

df %>%
  ggplot( aes(x=score_type, y=score)) +
  geom_boxplot(aes(color=sex),outlier.shape = NA, size=1.5, show.legend = F) +
  geom_point(aes(color=sex, shape = age_cat, group = sex),
 position=position_jitterdodge(dodge.width=0.9), size=3) +
  scale_color_manual(values=c("#0072B2", "#CC79A7"), name="",
 labels=c("Male", "Female")) +
  scale_shape_manual(name="", labels=c('Younger', 'Older'),
values=c(1, 17)) +
  theme_bw()+
  theme(panel.border = element_blank(), axis.ticks = element_blank(),
        legend.position=c(0.9, 0.65), legend.text=element_text(size=11),
 legend.title=element_text(size=11.5),
        panel.grid.major.x = element_blank() ,
        plot.title = element_text(size=11, face = "bold"),
 axis.title=element_text(size=13),
        axis.text.y = element_text(size=11),
        axis.text.x = element_text(size=11),
        plot.margin = unit(c(0.5,0.2,0,0.2), "cm")) +
  labs(title= "", x = "",y = "Score") +
  scale_y_continuous(breaks=c(0, 20, 40, 60, 80, 100),
labels=c('0', '20', '40', '60', '80', '100')) +
  expand_limits(x=5, y=70) +
  scale_x_discrete(labels = c("A", "B")) + 
  coord_cartesian(clip = "off")

edited Jan 12, 2022 at 23:59

answered Jan 12, 2022 at 23:09

Dion Groothof

1,4767 silver badges18 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

user7395965 Over a year ago

shape = age_cat in the aesthetics of geom_point is not enough, unfortunately. Data points from factor sex go outside the borders of the boxplots where they should belong... I much abbreviated my original data set for the question, so this is on easily visible.

Dion Groothof Over a year ago

I see, but you did not specify this in your question. I edited the answer accordingly. It should now match your aim. Cheers!

Collectives™ on Stack Overflow

ggplot boxplot with data points: how to control both point COLOR and SHAPE based on two factors?

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related