Rstudio crashes when using ggsave() to write .tiff images

[RuiyuRayWang]

When using ggsave() tiff device saving Seurat plots, Rstudio crashed:
R Session Aborted: R encountered a fatal error. The session was terminated.

Yesterday it was working fine. This only happens today.
This happens only on my mac (Catalina 10.15.7). Testing on ubuntu (18.04 LTS) everything works fine.

Searched for recently raised issues and found a similar one #4554 which is resolved by Rstudio upgrade. I tried upgrading Rstudio to 1.v1.4.1717-3 Preview but the issue persisted.

library(Seurat)
library(ggplot2)

DotPlot(pbmc_small, 
        features = "MS4A1")
ggsave(filename = "test.png", device = "png")  # Works fine
ggsave(filename = "test.tiff", device = "tiff") # Crash

Plot functions I've tested and crashed include: DotPlot(), RidgePlot(), VlnPlot(), FeaturePlot(), DimPlot(), FeatureScatter()

# All crashed when invoking ggsave("test.tiff")
DotPlot(pbmc_small, 
        features = "MS4A1")
RidgePlot(pbmc_small, 
        features = "MS4A1")
VlnPlot(pbmc_small,
        features = "MS4A1")
FeaturePlot(pbmc_small,
            features = "MS4A1")
DimPlot(pbmc_small,
        features = "MS4A1")
FeatureScatter(pbmc_small,
               feature1 = "MS4A1", feature2 = "CD79B")

DoHeatmap() is unaffected

library(tidyverse)
pbmc_small <- pbmc_small %>% NormalizeData() %>% ScaleData()
DoHeatmap(pbmc_small, features = VariableFeatures(pbmc_small))
ggsave(filename = "test.tiff", device = "tiff") # Success

This does not look like a ggplot issue as plots generated by ggplot are unaffected. Only Seurat generated plots are affected.

ggplot(mtcars, aes(x=mpg, y=cyl)) +
  geom_point()
ggsave("test.tiff") # Works fine

> sessionInfo()

R version 4.0.2 (2020-06-22)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Catalina 10.15.7

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

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

other attached packages:
[1] ggplot2_3.3.5      SeuratObject_4.0.2 Seurat_4.0.5      

loaded via a namespace (and not attached):
  [1] nlme_3.1-153          spatstat.sparse_2.0-0 matrixStats_0.61.0    RcppAnnoy_0.0.19      RColorBrewer_1.1-2   
  [6] httr_1.4.2            sctransform_0.3.2     tools_4.0.2           utf8_1.2.2            R6_2.5.1             
 [11] irlba_2.3.3           rpart_4.1-15          KernSmooth_2.23-20    uwot_0.1.10           mgcv_1.8-38          
 [16] DBI_1.1.1             lazyeval_0.2.2        colorspace_2.0-2      withr_2.4.2           tidyselect_1.1.1     
 [21] gridExtra_2.3         compiler_4.0.2        plotly_4.10.0         scales_1.1.1          lmtest_0.9-38        
 [26] spatstat.data_2.1-0   ggridges_0.5.3        pbapply_1.5-0         goftest_1.2-3         stringr_1.4.0        
 [31] digest_0.6.28         spatstat.utils_2.2-0  rmarkdown_2.11        pkgconfig_2.0.3       htmltools_0.5.2      
 [36] parallelly_1.28.1     fastmap_1.1.0         htmlwidgets_1.5.4     rlang_0.4.12          shiny_1.7.1          
 [41] generics_0.1.1        zoo_1.8-9             jsonlite_1.7.2        ica_1.0-2             dplyr_1.0.7          
 [46] magrittr_2.0.1        patchwork_1.1.1       Matrix_1.3-4          Rcpp_1.0.7            munsell_0.5.0        
 [51] fansi_0.5.0           abind_1.4-5           reticulate_1.22       lifecycle_1.0.1       stringi_1.7.5        
 [56] yaml_2.2.1            MASS_7.3-54           Rtsne_0.15            plyr_1.8.6            grid_4.0.2           
 [61] parallel_4.0.2        listenv_0.8.0         promises_1.2.0.1      ggrepel_0.9.1         crayon_1.4.2         
 [66] deldir_1.0-6          miniUI_0.1.1.1        lattice_0.20-45       cowplot_1.1.1         splines_4.0.2        
 [71] tensor_1.5            knitr_1.36            pillar_1.6.4          igraph_1.2.7          spatstat.geom_2.3-0  
 [76] future.apply_1.8.1    reshape2_1.4.4        codetools_0.2-18      leiden_0.3.9          glue_1.4.2           
 [81] evaluate_0.14         data.table_1.14.2     png_0.1-7             vctrs_0.3.8           httpuv_1.6.3         
 [86] polyclip_1.10-0       gtable_0.3.0          RANN_2.6.1            purrr_0.3.4           spatstat.core_2.3-1  
 [91] tidyr_1.1.4           scattermore_0.7       future_1.23.0         assertthat_0.2.1      xfun_0.27            
 [96] mime_0.12             xtable_1.8-4          later_1.3.0           survival_3.2-13       viridisLite_0.4.0    
[101] tibble_3.1.5          cluster_2.1.2         globals_0.14.0        fitdistrplus_1.1-6    ellipsis_0.3.2       
[106] ROCR_1.0-11

[RuiyuRayWang]

Update:

Encountered the same issue without using Seurat but using cowplot with ggpubr. I guess cowplot is the source of the issue here?

Also tried save_plot() to replace ggsave() but the issue persisted. Guess I'll close this issue and post it to cowplot page.

[wibeasley]

We had this problem too with a vanilla ggplot graph. We got around it with grDevices::tiff() and dev.off().

cc: @deshea

[RuiyuRayWang]

We had this problem too with a vanilla ggplot graph. We got around it with grDevices::tiff() and dev.off().

cc: @deshea

Thanks @wibeasley, grDevices works!

As pointed out by @clauswilke, this is an ragg issue. Switching to grDevices resolved the problem.

library(Seurat)
library(ggplot2)

DotPlot(pbmc_small, 
        features = "MS4A1")
ggsave("test.tiff", device = grDevices::tiff)

Alternatively

grDevices::tiff(filename = "test.tiff", bg = "transparent")
DotPlot(pbmc_small, 
        features = "MS4A1")
dev.off()