"Copy as metafile" in Rgui, as well as default ggsave(), scale the whole figure to the top left quarter

[telenskyt]

I have exactly the same problem as reported here, not sure why that one was closed.

When I produce ggplot2 and then right click the chart and select "Copy as metafile" (in the standard Rgui in Windows), it scales the whole figure just in the top left quarter like this:

1. In the Rgui, everything looks alright.
2. I get the same faulty result when I do ggsave("cover.wmf").

Reproducible code:

cover <- structure(list(scen = c(1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 
2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 5L, 5L, 
5L, 5L, 5L, 5L, 6L, 6L, 6L, 6L, 6L, 6L, 7L, 7L, 7L, 7L, 7L, 7L, 
8L, 8L, 8L, 8L, 8L, 8L, 9L, 9L, 9L, 9L, 9L, 9L, 10L, 10L, 10L, 
10L, 10L, 10L, 11L, 11L, 11L, 11L, 11L, 11L, 12L, 12L, 12L, 12L, 
12L, 12L, 13L, 13L, 13L, 13L, 13L, 13L, 14L, 14L, 14L, 14L, 14L, 
14L, 15L, 15L, 15L, 15L, 15L, 15L), n = c(20, 20, 20, 20, 20, 
20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 
20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 
20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 
20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 
20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 
20, 20, 20, 20, 20), r = c(0, 0, 0, 0, 0, 0, 0.2, 0.2, 0.2, 0.2, 
0.2, 0.2, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.6, 0.6, 0.6, 0.6, 0.6, 
0.6, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0, 0, 0, 0, 0, 0, 0.2, 0.2, 
0.2, 0.2, 0.2, 0.2, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.6, 0.6, 0.6, 
0.6, 0.6, 0.6, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0, 0, 0, 0, 0, 0, 
0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.4, 0.4, 0.4, 0.4, 0.4, 0.4, 0.6, 
0.6, 0.6, 0.6, 0.6, 0.6, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8), n_ind = c(40, 
40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 
40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 40, 160, 160, 
160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 
160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 160, 
160, 160, 640, 640, 640, 640, 640, 640, 640, 640, 640, 640, 640, 
640, 640, 640, 640, 640, 640, 640, 640, 640, 640, 640, 640, 640, 
640, 640, 640, 640, 640, 640), method = structure(c(3L, 4L, 5L, 
2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 
4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 
1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 
2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 
4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 1L, 3L, 4L, 5L, 2L, 6L, 
1L, 3L, 4L, 5L, 2L, 6L, 1L), levels = c("true", "ignore", "BPA2012", 
"corrected(1)", "corrected(100)", "model_est"), class = "factor"), 
    cover_count = c(3243L, 4916L, 4918L, 4772L, 4825L, 4769L, 
    3266L, 4918L, 4919L, 4751L, 4810L, 4774L, 3341L, 4886L, 4886L, 
    4733L, 4821L, 4751L, 3134L, 4798L, 4794L, 4687L, 4816L, 4783L, 
    2030L, 4285L, 4279L, 4298L, 4828L, 4745L, 1859L, 4810L, 4809L, 
    4733L, 4805L, 4722L, 1896L, 4822L, 4820L, 4739L, 4817L, 4747L, 
    2041L, 4821L, 4825L, 4744L, 4803L, 4737L, 2284L, 4827L, 4826L, 
    4752L, 4815L, 4756L, 2760L, 4806L, 4808L, 4773L, 4835L, 4758L, 
    959L, 4740L, 4739L, 4719L, 4796L, 4721L, 976L, 4754L, 4761L, 
    4738L, 4805L, 4734L, 1101L, 4755L, 4759L, 4732L, 4793L, 4728L, 
    1202L, 4795L, 4796L, 4772L, 4820L, 4765L, 1610L, 4842L, 4840L, 
    4788L, 4835L, 4758L), cover = c(0.6486, 0.9832, 0.9836, 0.9544, 
    0.965, 0.9538, 0.6532, 0.9836, 0.9838, 0.9502, 0.962, 0.9548, 
    0.6682, 0.9772, 0.9772, 0.9466, 0.9642, 0.9502, 0.6268, 0.9596, 
    0.9588, 0.9374, 0.9632, 0.9566, 0.406, 0.857, 0.8558, 0.8596, 
    0.9656, 0.949, 0.3718, 0.962, 0.9618, 0.9466, 0.961, 0.9444, 
    0.3792, 0.9644, 0.964, 0.9478, 0.9634, 0.9494, 0.4082, 0.9642, 
    0.965, 0.9488, 0.9606, 0.9474, 0.4568, 0.9654, 0.9652, 0.9504, 
    0.963, 0.9512, 0.552, 0.9612, 0.9616, 0.9546, 0.967, 0.9516, 
    0.1918, 0.948, 0.9478, 0.9438, 0.9592, 0.9442, 0.1952, 0.9508, 
    0.9522, 0.9476, 0.961, 0.9468, 0.2202, 0.951, 0.9518, 0.9464, 
    0.9586, 0.9456, 0.2404, 0.959, 0.9592, 0.9544, 0.964, 0.953, 
    0.322, 0.9684, 0.968, 0.9576, 0.967, 0.9516)), row.names = c(NA, 
-90L), class = "data.frame")


library(ggplot2)

ggplot(cover, aes(x = r, y = cover)) + ylab("coverage (95% CI)") + 
	geom_hline(yintercept = 0.95) + 
	geom_point(color = "#00B937")

ggplot2 version 3.5.1 in R 4.4.1

[teunbrand]

Thanks for the report! I cannot reproduce this on my end on Windows 11 R 4.4.2 and ggplot2 3.5.1. Opening the .wmf file in paint or affinity designer just shows the plot panel filling all the appropriate area. Using the 'copy as metafile' option in the RGui seems normal too.
What happens on your end if you repeat the procedure on the following graphic?

library(grid)
grid.newpage()
grid.draw(rectGrob(width = 0.9, height = 0.9))

[telenskyt]

Thanks @teunbrand. Your code produces the same faulty behaviour on my setup.

It is however true that there might perhaps be some non-determinism in this. Because I observed the issue a month ago with different code, and not with this one, after I restarted the R session, now I observe it with this one.

[teunbrand]

Plots and the grid based rectangle rectangle above should dynamically adapt to fit the available size reported by the device during rendering. Based on #4541 (comment), you might try to set ggsave("cover.wmf", xpinch = 96, ypinch = 96). Other than this, I'm afraid it is more of a graphics device issue than a ggplot2 issue and mostly out of our hands.

[telenskyt]

Thanks @teunbrand.

1. The ggsave() workaround doesn't work for me.
2. You are right that this bug is also present without ggplot2, e.g. with plot(1, 1)!

[teunbrand]

Thanks for confirming with plot()! In this case, I'll close this issue as there is nothing actionable that ggplot2 can do here.