-
Notifications
You must be signed in to change notification settings - Fork 2.1k
Description
I'm observing that changing size
in geom_sf()
no longer affects linewidth:
library("ggplot2")
library("sf")
#> Linking to GEOS 3.9.1, GDAL 3.3.2, PROJ 7.2.1; sf_use_s2() is TRUE
nc <- sf::st_read(system.file("shape/nc.shp", package = "sf"), quiet = TRUE)
indices <- which(nc$CNTY_ID >= 1820 & nc$CNTY_ID <= 1830)
nc <- nc[indices, ]
p <- ggplot(nc, aes(fill = NAME)) +
theme_bw(15)
p + geom_sf(size = 3)
Changing linewidth
does affect linewidth:
p + geom_sf(linewidth = 3)
This breaking change isn't documented in the NEWS.md file and according to the discussion in #3672 it doesn't seem like such a "breaking change" was intended until after proper deprecation?:
@yutannihilation wrote:
Just curious, will this eventually change the behavior of size on Geoms that actually have both size and linewidth? I know there's no breaking change here, but #4872 made me think this might need some consideration to avoid ambiguity.
For example, in the case of geom_pointrange(), it's not clear what to do when we want to increase only the size of the points. Does specifying only size achieve this? Or, will this increase both the size of the points and that of the lines? I feel the former behavior is right, but it would be a breaking change.
and @thomasp85 replied:
At this point the ambiguous geoms will not differentiate but I think we would do that down the line after proper deprecation