sf + ggplot + ragg doesn't print degree symbol

[mhpob]

library(sf)

nc <- sf::st_read(system.file("shape/nc.shp", package="sf"))


library(ggplot2)
map <- ggplot() +
    geom_sf(data = nc)


ragg::agg_png(width = 900, height = 450)
map
invisible(dev.off())

png(width = 900, height = 450)
map
invisible(dev.off())

[thomasp85]

I cannot reproduce this. What system are you on? And does the issue persist with the development version?

[mhpob]

Using Windows 10 -- just re-installed the development version of ragg and the problem persists. Here's my sessionInfo:

R version 4.0.2 (2020-06-22)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19041)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252 LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] ragg_0.3.1.9000 ggplot2_3.3.2 sf_0.9-6

loaded via a namespace (and not attached):
[1] Rcpp_1.0.5 rstudioapi_0.11 magrittr_1.5 units_0.6-7 munsell_0.5.0 tidyselect_1.1.0 colorspace_1.4-1 R6_2.4.1
[9] rlang_0.4.7 dplyr_1.0.2 tools_4.0.2 grid_4.0.2 gtable_0.3.0 KernSmooth_2.23-17 e1071_1.7-3 DBI_1.1.0
[17] withr_2.3.0 systemfonts_0.3.2 ellipsis_0.3.1 class_7.3-17 tibble_3.0.3 lifecycle_0.2.0 crayon_1.3.4 textshaping_0.1.1
[25] farver_2.0.3 purrr_0.3.4 vctrs_0.3.4 glue_1.4.2 compiler_4.0.2 pillar_1.4.6 scales_1.1.1 generics_0.0.2

[thomasp85]

Thanks - I'll try to check this out on windows.

[thomasp85]

I can reproduce on Windows

[thomasp85]

OK, so apparently the symbol font does not provide a Unicode character map but only understands the ancient Adobe Symbol encoding... this is fixable but I'll have to think about the path of least resistance (esp since it is a windows only problem)

[MargaretSiple-NOAA]

I have had this same issue recently, though I can't pinpoint when it started happening. When I use ggplot() + geom_sf(), degree symbols appear as little boxes, as they did in @mhpob 's example above. When I change to R 4.2.1 instead of 4.2.2, it resolves the issue. I have noooo idea what is happening.

Here is a reprex that will reproduce this same issue on my Windows 10 machine:

# Map test based on geom_sf() help file
nc <- sf::st_read(system.file("shape/nc.shp", package = "sf"), quiet = TRUE)
ggplot(nc) +
  geom_sf(aes(fill = AREA))

which produces the following:

I suspect that this is related to my windows fonts somehow, but any additional insight would be appreciated. I will struggle mightily before developing a hacky workaround. Thank you!

Here is my sessionInfo():

R version 4.2.2 (2022-10-31 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19044)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.utf8 LC_CTYPE=English_United States.utf8
[3] LC_MONETARY=English_United States.utf8 LC_NUMERIC=C
[5] LC_TIME=English_United States.utf8

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] ggplot2_3.4.0

loaded via a namespace (and not attached):
[1] Rcpp_1.0.9 pillar_1.8.1 compiler_4.2.2
[4] class_7.3-20 R.methodsS3_1.8.2 R.utils_2.12.2
[7] tools_4.2.2 digest_0.6.30 lattice_0.20-45
[10] lifecycle_1.0.3 tibble_3.1.8 gtable_0.3.1
[13] R.cache_0.16.0 pkgconfig_2.0.3 rlang_1.0.6
[16] DBI_1.1.3 cli_3.4.1 rstudioapi_0.14
[19] rnaturalearthdata_0.1.0 e1071_1.7-12 withr_2.5.0
[22] dplyr_1.0.10 styler_1.8.1 generics_0.1.3
[25] vctrs_0.5.0 classInt_0.4-8 grid_4.2.2
[28] tidyselect_1.2.0 glue_1.6.2 sf_1.0-9
[31] R6_2.5.1 rnaturalearth_0.1.0 fansi_1.0.3
[34] sp_1.5-1 farver_2.1.1 purrr_0.3.5
[37] magrittr_2.0.3 units_0.8-0 scales_1.2.1
[40] rsconnect_0.8.28 assertthat_0.2.1 colorspace_2.0-3
[43] labeling_0.4.2 KernSmooth_2.23-20 utf8_1.2.2
[46] proxy_0.4-27 munsell_0.5.0 R.oo_1.25.0

[grwhumphries]

@MargaretSiple-NOAA - Ditto here - having the exact same problem with the Degree symbols not rendering.

[MargaretSiple-NOAA]

@grwhumphries Does it resolve if you use R 4.2.1 or 4.2.0 instead of 4.2.2? I am not sure if the version is the issue but that's how I was able to resolve the issue temporarily. Not ideal, but it did work!

[sraul1]

Just updated to 4.2.2 and have this issue as well. Resolves itself in 4.2.1. here is my sessionInfo() (sorry for all the packages loaded, too deep into something to start a new session, I'll update with a cleaner environment soon)

R version 4.2.2 (2022-10-31 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19044)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.utf8  LC_CTYPE=English_United States.utf8   
[3] LC_MONETARY=English_United States.utf8 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.utf8    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] reticulate_1.26 lubridate_1.8.0 tidyterra_0.3.0 terra_1.6-17    readxl_1.4.1    broom_1.0.1    
 [7] ggpubr_0.5.0    forcats_0.5.2   stringr_1.4.1   dplyr_1.0.10    purrr_0.3.5     readr_2.1.3    
[13] tidyr_1.2.1     tibble_3.1.8    ggplot2_3.4.0   tidyverse_1.3.2

loaded via a namespace (and not attached):
 [1] httr_1.4.4          jsonlite_1.8.3      splines_4.2.2       carData_3.0-5       modelr_0.1.9       
 [6] assertthat_0.2.1    googlesheets4_1.0.1 cellranger_1.1.0    yaml_2.3.6          pillar_1.8.1       
[11] backports_1.4.1     lattice_0.20-45     glue_1.6.2          digest_0.6.30       ggsignif_0.6.4     
[16] rvest_1.0.3         colorspace_2.0-3    htmltools_0.5.3     Matrix_1.5-1        pkgconfig_2.0.3    
[21] haven_2.5.1         s2_1.1.0            scales_1.2.1        tzdb_0.3.0          proxy_0.4-27       
[26] googledrive_2.0.0   mgcv_1.8-41         generics_0.1.3      farver_2.1.1        car_3.1-1          
[31] ellipsis_0.3.2      withr_2.5.0         cli_3.4.1           magrittr_2.0.3      crayon_1.5.2       
[36] evaluate_0.18       fs_1.5.2            fansi_1.0.3         nlme_3.1-160        class_7.3-20       
[41] rstatix_0.7.1       xml2_1.3.3          data.table_1.14.4   tools_4.2.2         hms_1.1.2          
[46] gargle_1.2.1        lifecycle_1.0.3     munsell_0.5.0       reprex_2.0.2        compiler_4.2.2     
[51] e1071_1.7-12        rlang_1.0.6         units_0.8-0         classInt_0.4-8      grid_4.2.2         
[56] rstudioapi_0.14     labeling_0.4.2      rmarkdown_2.17      wk_0.7.0            gtable_0.3.1       
[61] codetools_0.2-18    abind_1.4-5         DBI_1.1.3           polynom_1.4-1       R6_2.5.1           
[66] knitr_1.40          fastmap_1.1.0       utf8_1.2.2          KernSmooth_2.23-20  stringi_1.7.8      
[71] Rcpp_1.0.9          png_0.1-7           vctrs_0.5.0         sf_1.0-8            dbplyr_2.2.1       
[76] tidyselect_1.2.0    xfun_0.34         

[icrichmond]

Also having this issue on a Windows system since updating to R 4.2.2 (resolved once I switched to R 4.2.1). Here is my sessionInfo() (sorry if its messy, this is my first time attaching sessionInfo!)

Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19044)

Matrix products: default

locale:
[1] LC_COLLATE=English_Canada.utf8  LC_CTYPE=English_Canada.utf8   
[3] LC_MONETARY=English_Canada.utf8 LC_NUMERIC=C                   
[5] LC_TIME=English_Canada.utf8    

attached base packages:
[1] stats     graphics  grDevices datasets  utils     methods   base     

other attached packages:
[1] ggplot2_3.4.0   sf_1.0-9        dplyr_1.0.10    targets_0.14.0  patchwork_1.1.2

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.9          pillar_1.8.1        compiler_4.2.2      class_7.3-20       
 [5] tools_4.2.2         qs_0.25.4           digest_0.6.30       lifecycle_1.0.3    
 [9] tibble_3.1.8        gtable_0.3.1        pkgconfig_2.0.3     rlang_1.0.6        
[13] igraph_1.3.5        DBI_1.1.3           cli_3.4.1           ggrepel_0.9.2      
[17] yaml_2.3.6          xfun_0.35           e1071_1.7-12        withr_2.5.0        
[21] knitr_1.41          generics_0.1.3      vctrs_0.5.1         classInt_0.4-8     
[25] grid_4.2.2          tidyselect_1.2.0    RApiSerialize_0.1.2 glue_1.6.2         
[29] data.table_1.14.6   R6_2.5.1            processx_3.8.0      fansi_1.0.3        
[33] base64url_1.4       farver_2.1.1        callr_3.7.3         magrittr_2.0.3     
[37] units_0.8-0         backports_1.4.1     scales_1.2.1        ps_1.7.2           
[41] codetools_0.2-18    colorspace_2.0-3    renv_0.16.0         KernSmooth_2.23-20 
[45] stringfish_0.15.7   utf8_1.2.2          proxy_0.4-27        RcppParallel_5.1.5 
[49] munsell_0.5.0 ```

[grwhumphries]

I switched back to 4.2.1 and the issue was resolved. Unsure what's going on there. Hopefully gets picked up

[thomasp85]

Thanks for the reports everyone... I'll see if I can figure out what is going on - weird how the R version can affect this

[dempsey-CMAR]

Hi folks! I just updated to R 4.2.2 and am also having this problem.

Session Info:

R version 4.2.2 (2022-10-31 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 18363)

Matrix products: default

locale:
[1] LC_COLLATE=English_Canada.utf8  LC_CTYPE=English_Canada.utf8    LC_MONETARY=English_Canada.utf8
[4] LC_NUMERIC=C                    LC_TIME=English_Canada.utf8    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

loaded via a namespace (and not attached):
[1] compiler_4.2.2  tools_4.2.2     rstudioapi_0.14

[fpetersma]

Hi all! I had the same issue in R 4.2.2, and it got resolved when I switched back to R 4.2.1.

[thomasp85]

I'm looking into it... will hopefully have a fix soon

[thomasp85]

Hmm... I have now tried to reproduce this on a Widows machine. I am completely unable to reproduce this issue in ragg. However, I do get the reported behaviour with the default windows device so I suspect that this is the device you are all using.

To make sure, can I get any of the above posters to run this piece of code and report if they indeed experience the issue:

library(grid)
ragg::agg_png("unicode_test.png")
grid.newpage()
grid.text(expression(90*degree))
dev.off()

[MargaretSiple-NOAA]

Hi @thomasp85 -- I ran the code above (now using R 4.2.2) and the piece of code returned this. It seems like the issue doesn't occur with the code you gave, but I didn't know I had changed the device I was using! Not sure what else it could be.

When I run the code I pasted in my example above, I still get the sad squares-instead-of-degrees issue. I'm on a Windows machine too. I hope this gets us one step closer to solving the mystery...

Here's the output from your code scrap:

[thomasp85]

Thanks for responding so quickly @MargaretSiple-NOAA — can you be more specific about how you generate the plot where you see the issue? Do you open a device with ragg::agg_png() and plot it? Do you see it in the plot pane in RStudio or in the graphics window in the R gui.

The output of my provided code heavily suggests you are not using ragg in any way

[dempsey-CMAR]

Thanks for looking into this @thomasp85.

I can see the degree symbol when I run your code.

I don't think I am using ragg to make my maps (unless it is under the hood somewhere).

Based on your comment above, I realized that the issue depends on how I am viewing the plot.

Here is a quick reprex:

library(rnaturalearth)
library(ggplot2)

can <- ne_countries(country = "canada", returnclass = "sf")

ggplot() + geom_sf(data = can)

When I run the above code, this is what I see in the RStudio plot pane and in word and html documents generated from RMarkdown:

When I save the map as a png or jpg with ggsave OR render a pdf document with RMarkdown, the degree symbols show up.

This is my session info:

R version 4.2.2 (2022-10-31 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 18363)

Matrix products: default

locale:
[1] LC_COLLATE=English_Canada.utf8  LC_CTYPE=English_Canada.utf8   
[3] LC_MONETARY=English_Canada.utf8 LC_NUMERIC=C                   
[5] LC_TIME=English_Canada.utf8    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] sf_1.0-9            rnaturalearth_0.1.0 ggplot2_3.4.0      

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.9         pillar_1.8.1       compiler_4.2.2     class_7.3-20      
 [5] tools_4.2.2        lattice_0.20-45    lifecycle_1.0.3    tibble_3.1.8      
 [9] gtable_0.3.1       pkgconfig_2.0.3    rlang_1.0.6        DBI_1.1.3         
[13] cli_3.4.1          rstudioapi_0.14    e1071_1.7-12       withr_2.5.0       
[17] dplyr_1.0.10       generics_0.1.3     vctrs_0.5.1        classInt_0.4-8    
[21] grid_4.2.2         tidyselect_1.2.0   glue_1.6.2         R6_2.5.1          
[25] fansi_1.0.3        sp_1.5-1           farver_2.1.1       magrittr_2.0.3    
[29] scales_1.2.1       units_0.8-1        assertthat_0.2.1   colorspace_2.0-3  
[33] labeling_0.4.2     utf8_1.2.2         KernSmooth_2.23-20 proxy_0.4-27      
[37] munsell_0.5.0   

[thomasp85]

I'm pretty sure my ideas are right - all of your issues are related to the default windows device and not anything in ragg/ggplot2/sf

In general, the default windows graphics device is horrible and you should all avoid using it :-)

Going to close this. Please open another issue if an example that clearly shows an issue in ragg is observed

[dempsey-CMAR]

Good to know!
Any other advice on how to fix the issue? It didn't happen until I started using R 4.2.2, so maybe I just need to use an earlier version.
I use RMarkdown to generate reports with maps in Word for work, so would be nice to know how to fix!

[thomasp85]

in the RMarkdown file you have to specify a different graphics device than the default, using the dev option in the code chunk. See https://ragg.r-lib.org/#use-ragg-with-knitr for how to set it to using ragg

[dempsey-CMAR]

I will check that out - thanks again!

Edit - adding knitr::opts_chunk$set(dev = "ragg_png") to my RMarkdown document fixed the issue for me.

[MargaretSiple-NOAA]

@thomasp85 Thanks for that; I didn't realize that the default windows device was bad! I am grateful for this education. I changed my graphics device backend to AGG and also set it in my knitr options and all is well.

[phineas-pta]

anyone have a workaround in an interactive context? (meaning no ggsave, nor r markdown)

[thomasp85]

This is not an issue with ragg but a problem with the default png device shipped with R on windows. If you use RStudio you can set change the device used in the plot pane as per the instructions in the ragg read me