Closed larmarange closed 4 years ago
Yes. Thank you for this package.
I had the same issues here. I think linked to modifications in {rlang} and {tidyr}.
gather_
is now deprecated so that you need to use more tidyeval with gather
. I propose:
cols2 <- enquo(cols)
df <- tidyr::gather(data, "type", "value", !!cols2)
Concerning xvar
, I agree. I would do the same proposition:
xvar <- as.character(mapping$x)[2]
for changing from xvar <- as.character(mapping)['x']
to xvar <- as.character(mapping$x)[2]
, do you have any reproducible example?
> require(ggplot2)
Loading required package: ggplot2
> as.character(aes(x=abc))['x']
x
"abc"
> as.character(aes(x=abc)$x)
[1] "abc"
> as.character(aes(x=abc)$x)[2]
[1] NA
thanks @statnmap for the rlang
part.
I am using the dev version of {ggplot2}, this maybe a reason. Here are my outputs:
> as.character(aes(x=abc))['x']
[1] NA
> as.character(aes(x=abc)$x)
[1] "~" "abc"
> as.character(aes(x=abc)$x)[2]
[1] "abc"
>
To allow your function to work for both versions, you can add a test on xvar
like:
xvar <- as.character(aes(x=abc))['x']
if (is.na(xvar)) {
xvar <- as.character(mapping$x)[2]
}
thanks @statnmap and great to know the change.
Hello,
thank you for your package. When trying to use it, I have identified some bugs:
https://github.com/GuangchuangYu/scatterpie/blob/8a42783947a02ea45c55ac956f70dae559a27788/R/geom_scatterpie.R#L40 You should write
tidyr::gather_
instead of justgather
. Otherwise, you will have an error if the user did not chargetidyr
by himself.Similarly, the calls to
geom_arc_bar
should also be prefixed byggforce
.Finally, I had an issue with that line
https://github.com/GuangchuangYu/scatterpie/blob/9c8d2b262d16bd5f3352a5fd7e2c16229cf01611/R/geom_scatterpie.R#L31
xvar
was remaining empty.But, that alternative code was working:
xvar <- as.character(mapping$x)[2]
Best regards