英文:
Use tidy evaluation when passing an expression as a character to ggplot2::aes()
问题
我正在尝试将我在函数中使用的ggplot2转换为使用整洁评估，以避免评估的警告消息。特别是，我广泛使用了aes_string()，这些需要转换为aes()。当仅传递列的名称作为字符时，我可以处理这种情况。然而，我一直无法解决当字符是数学表达式时该如何处理的情况。
这里是一个小的可重现问题示例，我正在尝试解决这个问题。
library(ggplot2)
set.seed(1)
dat <- data.frame(x=rnorm(100), y=rnorm(100))
xvar <- 'x+100'
yvar <- 'y'
# 这个可以工作，但使用了不推荐使用的aes_string
ggplot(dat, aes_string(x=xvar, y=yvar))
# 这个可以工作
ggplot(dat, ggplot2::aes(x=x+100, y=.data[[yvar]]) + geom_point()
# 这个无法工作
ggplot(dat, aes(x={{xvar}}, y=.data[[yvar]]) + geom_point()
我的问题是，我需要使用哪些整洁评估技巧来使用xvar 来指定 x 变量，就像aes_string() 中那样可能？
英文:
I am trying to convert my use of ggplot2 in functions to using tidy evaluation so as to avoid the warning messages that are evaluated. I particular, I have extensively used aes_string() and these need to be converted to aes(). I can handle cases when just the name of a column is passed as a characater. However, I have been unable to work how to deal with the case when the character is a mathematical expression.
Here is a small reproducible example of the problem that I an trying to solve.
library(ggplot2)
set.seed(1)
dat <- data.frame(x=rnorm(100),y=rnorm(100))
xvar <- 'x+100'
yvar <- 'y'
#This works but uses the deprecated aes_string
ggplot(dat,aes_string(x=xvar,y=yvar))
#This works
ggplot(dat, ggplot2::aes(x=x+100, y=.data[[yvar]])) + geom_point()
#This does not work
ggplot(dat, aes(x={{xvar}}, y=.data[[yvar]])) + geom_point()
My question is what tidy evaluation techniques do I need to employ to use xvar to specify the x variable as is possible with aes_string()?
答案1
得分: 0
可能需要 parse_expr/eval
library(ggplot2)
ggplot(dat, aes(x=eval(rlang::parse_expr(xvar)), y=.data[[yvar]])) +
geom_point() +
xlab(xvar)
-output
或者另一种选择是插值和执行 eval/parse
eval(parse(text = glue::glue("ggplot(dat, aes(x = {xvar}, y = {yvar}))",
"+ geom_point()")))
英文:
It may need parse_expr/eval
library(ggplot2)
ggplot(dat, aes(x=eval(rlang::parse_expr(xvar)), y=.data[[yvar]])) +
geom_point() +
xlab(xvar)
-output
Or another option would be to interpolate and do the eval/parse
eval(parse(text = glue::glue("ggplot(dat, aes(x = {xvar}, y = {yvar}))",
"+ geom_point()")))
</details>
# 答案2
**得分**: 0
你可以使用 `eval(str2expression())`：
library(ggplot2)
ggplot(dat, aes(x = eval(str2expression(xvar)), y = eval(str2expression(yvar)))) +
geom_point() +
labs(x = xvar, y = yvar)
或者使用类似的 rlang 函数：
library(rlang)
ggplot(dat, aes(x = eval_tidy(parse_expr(xvar)), y = eval_tidy(parse_expr(yvar)))) +
geom_point() +
labs(x = xvar, y = yvar)
请注意，你需要手动使用 `labs()` 来设置轴标签；否则，你会得到像 x 轴上的 `"eval(str2expression(xvar))"` 这样的标签。
<details>
<summary>英文:</summary>
You could use `eval(str2expression())`:
library(ggplot2)
ggplot(dat, aes(x = eval(str2expression(xvar)), y = eval(str2expression(yvar)))) +
geom_point() +
labs(x = xvar, y = yvar)
Or using the analogous rlang functions:
library(rlang)
ggplot(dat, aes(x = eval_tidy(parse_expr(xvar)), y = eval_tidy(parse_expr(yvar)))) +
geom_point() +
labs(x = xvar, y = yvar)
[![][1]][1]
Note you’ll want to manually set the axis labels using `labs()`; otherwise you’ll end up with e.g. `"eval(str2expression(xvar))"` for the x axis.
[1]: https://i.stack.imgur.com/TMmr4.jpg
</details>
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论