geom_label_repel(..., point.padding = NA) throws error: "'x' argument must be a unit object"

[mlell]

Summary

geom_text_repel(..., point.padding = NA) can be used to avoid text being repelled by other data points. I need this feature because I want to use ggrepel to annotate inside a bar plot, so the text must not be repelled by the bars.

However, if I use geom_label_repel instead in this way, the following error is produced:

Error in convertUnit(x, unitTo, "x", "dimension", "x", "dimension", valueOnly = valueOnly) :
'x' argument must be a unit object

Since Issue #79 / PR #80 it is possible to use

Minimal code example

Here is the minimum amount of code neeeded to demonstrate the issue:

library(ggplot2)
library(ggrepel)
library(dplyr)
t <- tibble(Env = rep(LETTERS[1:3], c(4,3,2)), 
            nRpl = c(4:1, 3:1, 2:1),
            nGt = c(1, 2, 4, 50, 1, 2, 10, 5, 7))
t <- t %>% mutate(nRpl = as.factor(nRpl))
t <- t %>% group_by(Env) %>% mutate(lPos = cumsum(nGt)-nGt/2) # label positions
ggplot(t, aes(x = Env, y = nGt, fill = nRpl)) + geom_col() + 
     geom_text_repel(aes(y = lPos, label = nRpl), point.padding = NA)

Here is an image of the output produced by the code:

However, this does not work with geom_label_repel:

ggplot(t, aes(x = Env, y = nGt, fill = nRpl)) + geom_col() + 
       geom_label_repel(aes(y = lPos, label = nRpl), point.padding = NA)
## Error in convertUnit(x, unitTo, "x", "dimension", "x", "dimension", valueOnly = valueOnly) : 
##   'x' argument must be a unit object

Suggestions

It would be nice if geom_label_repel could also ignore other non-text geoms like is the case for geom_text_repel, for example if there are contrast issues between text and underlying graphics.

A similar topic was issue #79.

Version information

Here is the output from sessionInfo() in my R session:

R version 3.4.3 (2017-11-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS: /opt/Bio/R/3.4.3/lib64/R/lib/libRblas.so
LAPACK: /opt/Bio/R/3.4.3/lib64/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

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

other attached packages:
 [1] ggrepel_0.7.0     prPretty_0.0      bindrcpp_0.2      viridis_0.4.1    
 [5] viridisLite_0.2.0 tidyr_0.7.2       ggplot2_2.2.1     glue_1.2.0       
 [9] stringr_1.2.0     readxl_1.0.0      magrittr_1.5      dplyr_0.7.4      

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.15       RColorBrewer_1.1-2 cellranger_1.1.0  
 [4] pillar_1.1.0       compiler_3.4.3     plyr_1.8.4        
 [7] bindr_0.1          tools_3.4.3        digest_0.6.14     
[10] memoise_1.1.0      tibble_1.4.2       gtable_0.2.0      
[13] pkgconfig_2.0.1    rlang_0.1.6        cli_1.0.0         
[16] rstudioapi_0.7     commonmark_1.4     yaml_2.1.16       
[19] gridExtra_2.3      xml2_1.2.0         withr_2.1.1       
[22] knitr_1.18         roxygen2_6.0.1     devtools_1.13.4   
[25] rprojroot_1.3-2    grid_3.4.3         R6_2.2.2          
[28] purrr_0.2.4        backports_1.1.2    scales_0.5.0      
[31] assertthat_0.2.0   colorspace_1.3-2   labeling_0.3      
[34] utf8_1.1.3         stringi_1.1.6      lazyeval_0.2.1    
[37] munsell_0.4.3      crayon_1.3.4   

[slowkow]

Thanks for reporting this issue! Also, thank you so much for providing a runnable example!! 👍

It was a bug in my code, but it should be fixed now.

Here's an example with the development version of ggrepel:

library(ggplot2)
library(ggrepel)
library(dplyr)

t <- tibble(Env = rep(LETTERS[1:3], c(4,3,2)),
            nRpl = c(4:1, 3:1, 2:1),
            nGt = c(1, 2, 4, 50, 1, 2, 10, 5, 7))
t <- t %>% mutate(nRpl = as.factor(nRpl))
t <- t %>% group_by(Env) %>% mutate(lPos = cumsum(nGt)-nGt/2) # label positions

p <- ggplot(t, aes(x = Env, y = nGt, fill = nRpl)) + geom_col()

# Works
# p + geom_text_repel(aes(y = lPos, label = nRpl), point.padding = NA)

# Should work now!
p + geom_label_repel(aes(y = lPos, label = nRpl), point.padding = NA)