Really wrong code or qplot was not updated to the latest 'rlang' version?

[Rodrigo Botafogo]

Hello....

I have the following code:

    library('ggplot2')
    library('ISLR')
    library('MASS')
 
    err_plot = function(lstat, medv) {
         qplot(!!lstat, !!medv)
    }
 
    err_plot(Boston$lstat, Boston$medv)

And this issues the following error:

    Error: `expr` must quote a symbol, scalar, or call
    Call `rlang::last_error()` to see a backtrace
 

I thought this should be valid code in 'tidyverse' and the expr should 'inline ...' the given vectors.  Is the code illegal or is it that qplot was not updated?

Thanks!

[Marios Barlas]

What's the double.!! In your qplot call? Never seen it before

Get Outlook for Android

---

From: ggp...@googlegroups.com <ggp...@googlegroups.com> on behalf of Rodrigo Botafogo <rodrigo.a...@gmail.com>
Sent: Thursday, January 24, 2019 4:51:40 PM
To: ggplot2
Subject: Really wrong code or qplot was not updated to the latest 'rlang' version?

 

--
--
You received this message because you are subscribed to the ggplot2 mailing list.
Please provide a reproducible example: https://github.com/hadley/devtools/wiki/Reproducibility
 
To post: email ggp...@googlegroups.com
To unsubscribe: email ggplot2+u...@googlegroups.com
More options: http://groups.google.com/group/ggplot2

---
You received this message because you are subscribed to the Google Groups "ggplot2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ggplot2+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Brandon Hurr]

I think the error is saying that you’ve missed a step. 

!! (Bang bang) is expecting a symbol so you need to make your input variables symbols with  rlang::sym(). 

err_plot = function(lstat, medv) {

lstat <-  sym(lstat)

medv <- sym(medv)

qplot(!!lstat, !!medv)

}

My favorite tutorial for tidy evaluation is here:

https://edwinth.github.io/blog/dplyr-recipes/

I know that Hadley has also updated his Advanced R book. 

https://adv-r.hadley.nz/quasiquotation.html

HTH, 

B

--

[Roman Luštrik]

I thought qplot went by the way of the dodo?

Cheers,

Roman

--

In God we trust, all others bring data.

[Brandon Hurr]

I remembered something like this too, but I don't see anything in the recent release notes to indicate that this is the case. 

https://github.com/tidyverse/ggplot2/releases

Also, it does still work. 

library(ggplot2)

qplot(mtcars$hp, mtcars$disp)

B