Extract dplyr tbl and create vector where column 1 = column 2 - r

Question

In order to map colours to group, I am using the scale_colour_manual(values = c("G1" = "grey", ...)) function of the {ggplot2}package.
I have a main tibble with data where you can find groups, and I would like to highlight a specific group. Here G3 is highlighted, however this is not necessarily the case for all plots I want to generate.
Here is some sample data:
groups <- as_tibble(c("G1", "G2", "G3"))
colours <- as_tibble(c("grey", "grey", "purple"))
I then pull the vectors, but I don't know how to get the result mentioned above (values = c("G1" = "grey", ...))
groups_vec <- groups %>% pull()
colours_vec <- colours %>% pull()
myvalues <- c(groups_vec = colours_vec)
# this code returns the following
groups_vec1 groups_vec2 groups_vec3
"grey" "grey" "purple"
whereas I expect the following result:
c("G1" = "grey", "G2" = "grey", "G3" = "purple")
G1 G2 G3
"grey" "grey" "purple"
I can't find the right words to describe my problem, hope the example is clear enough.

Answer 1

The following should help.
library(dplyr)
library(ggplot2)
# simulate some data
some_data <- tibble(
GROUPS = c("Group1", "Group2", "Group3","Group4"),
VALUES = c(2,5,9,3),
COL = c("grey","grey","lightblue","grey")
)
# plot
ggplot(data = some_data) +
geom_col(aes(x = GROUPS, y = VALUES, fill = COL)) +
scale_fill_manual(values = c("grey", "green"))
Try to understand how the variable "COL" and the "values" in scale_fill_manual() work together.
EDITED ANSWER
You did not provide a reproducible example. Thus, the answer might not be exactly what you are fishing for. But I truly hope the following will help you understand how {ggplot} works with aesthetics and how you can control the way these aesthetics are presented.
I build the example on a simple geom_col(). You might have another use-case. But the principle should be transferable to all geoms. Just note that for some geoms you need to use color instead of fill.
P.S. I also recommend to use a single data-frame for your plot. There is no need to keep every variable in a separate tibble. Just add a column with the color-flag you want to use. This might simplify your code ... and the number of objects you use.
Let's start with a simple plot of what we have.
library(dplyr)
library(ggplot2)
# simulate some data
some_data <- tibble(
GROUPS = c("Group1", "Group2", "Group3","Group4"),
VALUES = c(2,5,9,3),
COL = c("grey","grey","purple","grey")
)
# understand how ggplot uses "categories" based on COL variable
# without specification ggplot uses the default colors for 2 different "categories", i.e. grey and purple
p <- ggplot(data = some_data) +
geom_col(aes(x = GROUPS, y = VALUES, fill = COL))
p
Please note that {ggplot} uses the default colors in a specific sequence (as they come). We come to this back later, as this is the sequence you will need to control. To make this more prominent, recode your colors, e.g. instead of purple set it to highlight (or yes,Y). The characters are only a "category" for {ggplot}. The actual value is of secondary order (while we humans assign meaning to something like "green" or "purple").
This yields:
Next look into assigning colors. Key take away, the sequence of your color specification matters!
p1 <- p +
scale_fill_manual(values = c("grey", "purple")) +
labs(subtitle = "order of categories [grey, purple]")
p2 <- p +
scale_fill_manual(values = c("purple", "grey")) +
labs(subtitle = "order of categories [purple, grey]")
library(patchwork) # for demo purposes we plot both graphs side by side
p1 + p2
You did not provide a reproducible example. Thus, let's emulate your function f() by assigning a new color based on a value. (Your function might be more complex, but I understand it will give you a flag for the color).
# some operation that can happen in a function to change the color coding
# e.g. we pick the value 5
some_data2 <- some_data %>%
mutate(COL = case_when(
between(VALUES, 4,6) ~ "purple"
,TRUE ~ "grey"
))
# note - now we have a vector of values for each "row" (aka group) in your dataframe
color_vec <- c("grey", "yellow", "grey", "purple")
some_data2 <- some_data2 %>% mutate(COL2 = color_vec)
Have a look at the tibble:
some_data2
# A tibble: 4 × 4
GROUPS VALUES COL COL2
<chr> <dbl> <chr> <chr>
1 Group1 2 grey grey
2 Group2 5 purple yellow
3 Group3 9 grey grey
4 Group4 3 grey purple
Let's plot this tibble using aesthetic fill for column COL and set our color-vector as the desired sequence in scale_fill_manual():
some_data2 %>%
ggplot() +
geom_col(aes(x = GROUPS, y = VALUES, fill = COL)) +
scale_fill_manual(values = color_vec)
OOOOOPS!?! what happened here?
Again. try to understand color assigment and number of colors in your plot (aka "categories").
Let's now use our "new" color column, i.e. COL2.
some_data2 %>%
ggplot() +
geom_col(aes(x = GROUPS, y = VALUES, fill = COL2)) +
scale_fill_manual(values = color_vec)
Try to work out why this also not works.
What you probably want is a sequence of colors dependent on the values you want to highlight. Your example does not explicitly mention how you construct this. Above provides you a pointer how to handle this with a simple highlight vs no-highlight (2 categories). But you can obviously define more colors like in the following example. In the following example, we make use of the fact that our column COL2 uses the target colors based on your function. Please note that you could define the categories (aka breaks different from the color values - think factor label and level).
# note - we define values for the different "categories" that we expect
# in this example now 3 (order matters - c.f. above!)
categories <- some_data2$COL2
color_vec2 <- some_data2$COL2 # if you use flags that are different from colors you can define them here
some_data2 %>%
ggplot() +
geom_col(aes(x = GROUPS, y = VALUES, fill = COL2)) +
scale_fill_manual(
# breaks sets the sequence of your categories
breaks = categories
# values are the colors you want to use
, values = color_vec2)
You can use - for geom_col() - the fill aesthetic to have it colored in - what I call here - "categories".
With scale_fill_manual(), you control the color sequence of these categories. You may want to create this vector based on the order of colors.