R ggplot: colour and fill arguments-CodePudding

I am new in R and just started to learn ggplot. I am so confused by the syntax, I thought the "color" and "fill" arguments should always follow color names or RGB specifications. But I've seen many cases where "color" and "fill" in aes() were assigned with variables, see the below example.

ggplot(faithfuld, aes(waiting, eruptions))  
  geom_raster(aes(fill = density))

I couldn't find an explanation of such use in [R documentation][1]. What does it mean? coloring by factor/grouping? if fill and color are assigned with variables, where should colors be specified? in scale_colour_manual?

Besides, I noticed that if specifying colors and/or transparency in aes(), the specified colors or transparency won't realise. For instance, in the below code, alpha = 0.3 is not working, I can change the alpha to any value, and the transparency will always be 0.5 in plotting. Why is that?

Also, I noticed that if I deleted fill or alpha in the aex(), the following "scale_fill_manual" wouldn't work. So is it true that "scale_fill_manual" is dependent on the geom_xx()?

p <- ggplot(dfcc)   geom_ribbon(aes(x = yr, ymax = ciupper, ymin = cilower, fill = "", alpha = 0.3))  
  scale_fill_manual(values = "blue", labels = "CI95%")

Sorry for so many questions, I am just so confused, and any help will be appreciated! [1]:

2. Use `Species` as color:

iris |> 
  ggplot(aes(x = Sepal.Length, y = Sepal.Width))  
  geom_point(aes(color = Species))  
  theme_minimal()  
  theme(panel.grid = element_blank())

You get:

3. Modify `Species` with the color of your choice:

iris |> 
  ggplot(aes(x = Sepal.Length, y = Sepal.Width))  
  geom_point(aes(color = Species))  
  theme_minimal()  
  theme(panel.grid = element_blank())  
  scale_color_manual(values = c("red", "yellow", "purple"))

You get:

Hope this is helpful.

2. Use Species as color:

3. Modify Species with the color of your choice:

2. Use `Species` as color:

3. Modify `Species` with the color of your choice: