如何使用汽车包版本 3 在 scatterplot() 中使用图例参数

Question

I am new to R and trying to work through some examples in the excellent "R in Action" by Robert Kabacoff.

He makes use of the car package to make a scatterplot. While trying the following example, I found that I was getting warnings and the plot did not match the one in the book:

library(car)

scatterplot(mpg ~ wt | cyl, data=mtcars, lwd=2, span=0.75, 
            main="Scatter Plot of MPG vs. Weight by # Cylinders", 
            xlab="Weight of Car (lbs / 1000)", 
            ylab="Miles Per Gallon", legend.plot=TRUE, id.method="identify", 
            labels=row.names(mtcars), boxplot="xy")

After some digging, I discovered that the car package has updated to version 3, and the book uses version 2. Apparently, the argument input to most car functions has completely changed.

I did find documentation for it at https://cran.r-project.org/web/packages/car/car.pdf , but I couldn't completely follow how all the arguments work.

I got very close to getting the plot to work correctly with the following code:

scatterplot(mpg ~ wt | cyl, data = mtcars, pch=c(1,2,3), 
            smooth = list(smoother=loessLine, span = .75, lty.smooth=1), 
            main = "Scatter Plot of MPG vs. Weight by # Cylinders", 
            xlab = "Weight of Car (lbs/1000)", ylab = "Miles per Gallon", 
            legend = c(title="cyl", coords="topleft"), 
            id = list(method="identify"),
            showlabels = names(row.names(mtcars)), regLine=c(method=lm, lty=1))

But I can't seem to completely follow how to use the new arguments, especially the legend. I can't seem to be able to plot the legend by coordinates, it only seems to work if I use the coords="topleft" or coords="bottom" type arguments.

Can anyone explain how to use the legend argument in scatterplot() in the car package version 3? Specifically how to plot it in specific coordinates, and outside the plot area?

If anyone can point me to a tutorial that isn't version 2, that would be helpful as well.

Answer 1

In general, I find ggplot a lot easier than the base plot functions. You do need to set the cyl variable as a factor though.

ggplot(data = mtcars, aes(x = wt, y = mpg, color = as.factor(cyl), shape = as.factor(cyl))) +
 geom_point() + # plots the scatter plot
 geom_smooth(method = "lm", se = F) + # plots the linear model
 geom_smooth(se = F) + # plots the loess model
 theme_minimal() # changes some of the formatting

I also highly recommend R for Data Science by Hadley Wickham ( http://r4ds.had.co.nz/ ) for getting started in R. And its free!