different shapes for jittered points in box plots using R and ggplot2

Question

I have a box plot with jittered data points that looks like the following. The x-axis has three different regions and y-axis has salaries.

I used the following piece of code to plot.

 ggplot(df1, aes(x = Region, y = Income, fill = Sex, color = Sex), size = 10) + geom_boxplot(outlier.shape = NA)+ geom_point(position = position_jitterdodge(0.5), alpha=0.3)

One additional thing I am trying to do is to add different shapes for data points based on different ethnicities. For example, both males and females are divided into caucasian and african-american groups. I have a column Ethnicity in my data.table. I would like to make the data points corresponding to caucasian to trinagle-shaped and african-american to circle in the plot.

The data.table looks like this

   Region Income Sex    Ethnicity       
   <chr>   <dbl> <chr>  <chr>           
 1 Area1    2000 male   African-american
 2 Area1    3000 female African-american
 3 Area1    2000 male   African-american
 4 Area1    4000 male   African-american
 5 Area1   40050 female African-american
 6 Area1   60000 male   African-american
 7 Area1    2000 male   Caucasian       
 8 Area1    4562 female Caucasian       
 9 Area1    4568 male   Caucasian       
10 Area1    6573 male   Caucasian

 
   

I am just wondering if it is possible in the box plot. I just tried with the following, but it does not work.

 ggplot(df1, aes(x = Region, y = Income, fill = Sex, color = Sex, shape = Ethnicity), size = 10)

It looks like this now

I would appreciate any help.

Answer 1

Add ethnicity to the aes() of geom_point instead of where you have it.

df %>%
    ggplot(aes(x = Region, y = Income, fill = Sex, color = Sex), size = 10) +
    geom_boxplot(outlier.shape = NA)+
    geom_point(aes(shape = Ethnicity), position = position_jitterdodge(0.5), alpha=0.3)