在 R data.table 中按组绘制点

Question

I am trying to plot points using a by argument in R data.table - the plan is to colour the points by the grouping variable, but in development I noticed a behaviour I thought was strange. With data.table, the operation in j (of DT[i, j, by] ) is supposed to be carried out per level in the by , eg

library(data.table)
dtcars <- copy(mtcars)
setDT(dtcars)
dtcars[, mean(mpg), by=cyl]

But I am now trying to get this to plot points for each level of cyl separately. The black points are showing what data should be getting plotted in red, but it only seems to do those where cyl is 8 when I use the by

dtcars[, plot(mpg~hp, typ="n")]
dtcars[, points(mpg~hp, col="black")]
dtcars[, points(mpg~hp, col="red"), by=cyl]

Any idea what is going on, why it is only acting on the one value of cyl and how to have R plot points for all levels of cyl, with by? I've used data.table a lot and not noticed this behaviour before.

---

Bonus points if you can tell me how to return the index of the by value so I can index the colours giving the same effect as

dtcars[, points(mpg~hp, col=c("red", "blue", "green")[as.factor(cyl)])]

Becoming something like

dtcars[, points(mpg~hp, col=c("red", "blue", "green")[by_index]), by=cyl]

Answer 1

Would you consider using ggplot2 instead of base R for plotting? If so, try the following :

library(data.table)
library(ggplot2)

ggplot(mtcars, aes(x = hp, y = mpg, color = as.factor(cyl))) +
  geom_point() +
  scale_color_discrete(name = "cyl") +
  theme_linedraw()