Highlight a specific point in ggplot2

Question

I have a dataframe like this:

item <- c("item1", "item2", "item3", "item1", "item2", "item3", "item1", "item2", "item3")
group <- c("A", "B", "C", "A", "B", "C", "A", "B", "C")
level <- c(NA, NA, 40, NA, 25, NA, 30, NA, NA)

data <- cbind(item, group, level)
data <- as_tibble(data)
data <- type.convert(data, as.is = TRUE)

Which appears as follows:

item  group level

item1    A    NA
item2    B    NA
item3    C    40
item1    A    NA
item2    B    25
item3    C    NA
item1    A    30
item2    B    NA
item3    C    NA

Every item is univocally associated to a specific group ( item1 is always linked to group A, item2 always to group B , etc.)

To plot the graph of the data, I use this code:

   graph <- data %>%
     ggplot(aes(x=group, y=level)) +
     geom_point(colour="blue", size=3, na.rm=TRUE)

which shows this result:

Now, I would like to display in red the point with value 25 , selecting it by the item. I.e., "if the item2 (which corresponds to group B ) has a value != NA, display its value in red, keeping all the other values in blue".

I have thought to a if in a for-cycle , but I don't know if it is the right reasoning.

Thank you for helping!

CLOSED: SOLUTION FOUND

I have created a subset of the dataframe by item:

my_dot <- subset(data, item=="item2")
my_dot <- type.convert(my_dot, as.is = TRUE)

and added a line to the ggplot which load the subsetted dataframe "my_plot" in geom_point(data=my_dot, ...)

    graph <- data %>%
      ggplot(aes(x=group, y=level)) +
      geom_point(colour="blue", size=3, na.rm=TRUE) +
      geom_point(data=my_dot, aes(x=group, y=level), colour="red", size=5, na.rm=TRUE)

Here the result I was looking for:

Answer 1

This is a dupe, but for your data, a few options:

1. aes(color=(group == "B")), so we should always have two colors.

   ggplot(data, aes(x=group, y=level)) +
     geom_point(aes(color=(group == "B")), size=3, na.rm=TRUE) +
     scale_color_manual(values=c("blue", "red"))
   

   

2. aes(color=group), where we can specify different colors for each group. We'll name the color-vector:

   ggplot(data, aes(x=group, y=level)) +
     geom_point(aes(color=group), size=3, na.rm=TRUE) + 
     scale_color_manual(values=c(A="blue", B="red", C="blue"))
   

   

3. Overlay a new point over the old:

   ggplot(data, aes(x=group, y=level)) +
     geom_point(color="blue", size=3, na.rm=TRUE) +
     geom_point(color="red", size=3, na.rm=TRUE, data = ~subset(., group == "B"))
   

   

   (This is generally not the "canonical" method of ggplot2; note that it does not add a legend, because we've not assigned a field to the color aesthetic.)

   (The data= here uses ~ subset(.). The tilde ~ and dot . are important. You can use dplyr::filter(., ...) if you prefer.)

(For those with legends, the legend can be removed with guide=NULL or renamed with name="..", added to the scale_color_manual call.)