For my convenience, I want to write a function that adds a line plot and a scatter plot to an already existing plot. I defined:
addlinetoplot <- function(dataset, varx, vary)
{
p <- geom_line(data=dataset, aes_string(x=varx, y=vary)) +
geom_point(data=dataset, aes_string(x=varx, y=vary))
p
}
The above function does not work. It works if I remove the +
sign and the geom_point()
part. The above does not work because for some reason one cannot add those two geoms. The following example illustrates the problem.
Then, I tried:
df1 <- data.frame(c1 = c(1:10), c2 = c(1:10))
c1 <- c(1:10)
csq <- c1^2
df2 <- data.frame(c1 = c(1:10), csq)
pltbase <- ggplot() + geom_line(df1, aes(x="c1", y="c2"))
# This does not work.
pltbase + addlinetoplot(dataset=df2, varx = "c1", vary = "csq")
I figured the problem is with the statement:
addthistotheplot <- geom_line(data=df2, aes_string(x="c1", y="csq")) +
geom_point(data=df2, aes_string(x="c1", y="csq"))
Trying to define the above statement th开发者_如何学运维rows an error: non-numeric argument to binary operator.
How can I define the addlinetoplot()
function so that I can make:
pltbase + addlinetoplot(dataset=df2, varx = "c1", vary = "csq")
work. One way is to separate out the geom_line and geom_point and have different function definition for each. Is there a way to do it in the same function?
Thanks!
I can't find the reference right now, but there's an answer on SO somewhere that explains you can add to a ggplot plot by passing new arguments as a list. Accordingly, your function becomes:
addlinetoplot <- function(dataset, varx, vary) {
list(
geom_line(data=dataset, aes_string(x=varx, y=vary)),
geom_point(data=dataset, aes_string(x=varx, y=vary))
)
}
And then your plotting code looks like:
pltbase <- ggplot() + geom_line(data = df1, aes(x=c1, y=c2))
pltbase + addlinetoplot(df2, varx = "c1", vary = "csq")
Results in:
精彩评论