I am trying to create a function which allows column names to be passed in by the user and I find that the resulting plot is simply one dot in the middle of screen. What needs to change for it to work as intended? My minimalist example is:
library(ggplot2)
weather <- data.frame(Month = month.name[1:5], `Low Temperature` = 20:16,
`High Temperature` = seq(30, 22, -2), check.names = FALSE)
xVariable <- "Month"
yVariable <- "High Temperature"
yVariable <- enquo(yVariable)
ggplot(weather, aes(x = xVariable, y = !!yVariable)) geom_point()
CodePudding user response:
Because it's quoted text, instead of enquo
, use rlang::sym
xVariable <- "Month"
yVariable <- "High Temperature"
fun <- function(dat, xvar, yvar) {
xvar <- rlang::sym(xvar)
yvar <- rlang::sym(yvar)
p1 <-dat %>%
ggplot(aes(x = !!xvar, y = !!yvar))
geom_point()
return(p1)
}
fun(weather, xVariable, yVariable)
CodePudding user response:
You could use aes_string
instead of aes
.
library(ggplot2)
weather <- data.frame(Month = month.name[1:5], `Low Temperature` = 20:16,
`High Temperature` = seq(30, 22, -2), check.names = FALSE)
xVariable <- "Month"
yVariable <- "`High Temperature`"
ggplot(weather, aes_string(x = xVariable, y = yVariable)) geom_point()
CodePudding user response:
Note that enquo()
is only for function arguments.
If you have column names as strings, use the .data
pronoun:
fn <- function(data, x, y) {
data %>%
ggplot(aes(x = .data[[x]], y = .data[[y]]))
geom_point()
}
x <- "disp"
y <- "drat"
fn(mtcars, x, y)