Facetting with complex expressions does not skip layers with missing data

[mundl]

library(ggplot2)
set.seed(2804)
point <- data.frame(x = rnorm(n = 10),
                    y = rnorm(n = 10),
                    class = sample(x = 2, size = 10, replace = TRUE))

poly <- data.frame(x = rnorm(n = 4),
                   y = rnorm(n = 4))

Works as expected, points and polygons (same across facets) are drawn.

ggplot(point, aes(x = x, y = y)) +
    geom_point() +
    geom_polygon(data = poly, fill = NA, col = 1) +
    facet_wrap(vars(class))

Using any function inside vars() or simply vars((class)) gives an error.

ggplot(point, aes(x = x, y = y)) +                                                               
    geom_point() +                                                                                   
    geom_polygon(data = poly, fill = NA, col = 1) +                                                  
    facet_wrap(vars(as.character(class)))        
                                                    
#> Error in as.character(class): cannot coerce type 'builtin' to vector of type 'character'

Using the same code from above but not specifying a data argument in geom_polygon() works.

ggplot(point, aes(x = x, y = y)) +
    geom_point() +
    geom_polygon(fill = NA, col = 1) +
    facet_wrap(vars(as.character(class)))

Session info

devtools::session_info()
#> Session info -------------------------------------------------------------
#>  setting  value                       
#>  version  R version 3.5.0 (2018-04-23)
#>  system   x86_64, mingw32             
#>  ui       RTerm                       
#>  language (EN)                        
#>  collate  German_Austria.1252         
#>  tz       Europe/Berlin               
#>  date     2018-06-07
#> Packages -----------------------------------------------------------------
#>  package    * version    date       source                            
#>  assertthat   0.2.0      2017-04-11 CRAN (R 3.5.0)                    
#>  backports    1.1.2      2017-12-13 CRAN (R 3.5.0)                    
#>  base       * 3.5.0      2018-04-23 local                             
#>  bindr        0.1.1      2018-03-13 CRAN (R 3.5.0)                    
#>  bindrcpp     0.2.2      2018-03-29 CRAN (R 3.5.0)                    
#>  colorspace   1.3-2      2016-12-14 CRAN (R 3.5.0)                    
#>  compiler     3.5.0      2018-04-23 local                             
#>  curl         3.2        2018-03-28 CRAN (R 3.5.0)                    
#>  datasets   * 3.5.0      2018-04-23 local                             
#>  devtools     1.13.5     2018-02-18 CRAN (R 3.5.0)                    
#>  digest       0.6.15     2018-01-28 CRAN (R 3.5.0)                    
#>  dplyr        0.7.4      2017-09-28 CRAN (R 3.5.0)                    
#>  evaluate     0.10.1     2017-06-24 CRAN (R 3.5.0)                    
#>  ggplot2    * 2.2.1.9000 2018-06-07 Github (tidyverse/ggplot2@4db5122)
#>  glue         1.2.0      2017-10-29 CRAN (R 3.5.0)                    
#>  graphics   * 3.5.0      2018-04-23 local                             
#>  grDevices  * 3.5.0      2018-04-23 local                             
#>  grid         3.5.0      2018-04-23 local                             
#>  gtable       0.2.0      2016-02-26 CRAN (R 3.5.0)                    
#>  htmltools    0.3.6      2017-04-28 CRAN (R 3.5.0)                    
#>  httr         1.3.1      2017-08-20 CRAN (R 3.5.0)                    
#>  knitr        1.20       2018-02-20 CRAN (R 3.5.0)                    
#>  labeling     0.3        2014-08-23 CRAN (R 3.5.0)                    
#>  lazyeval     0.2.1      2017-10-29 CRAN (R 3.5.0)                    
#>  magrittr     1.5        2014-11-22 CRAN (R 3.5.0)                    
#>  memoise      1.1.0      2017-04-21 CRAN (R 3.5.0)                    
#>  methods    * 3.5.0      2018-04-23 local                             
#>  mime         0.5        2016-07-07 CRAN (R 3.5.0)                    
#>  munsell      0.4.3      2016-02-13 CRAN (R 3.5.0)                    
#>  pillar       1.2.2      2018-04-26 CRAN (R 3.5.0)                    
#>  pkgconfig    2.0.1      2017-03-21 CRAN (R 3.5.0)                    
#>  plyr         1.8.4      2016-06-08 CRAN (R 3.5.0)                    
#>  R6           2.2.2      2017-06-17 CRAN (R 3.5.0)                    
#>  Rcpp         0.12.16    2018-03-13 CRAN (R 3.5.0)                    
#>  rlang        0.2.1      2018-05-30 CRAN (R 3.5.0)                    
#>  rmarkdown    1.9        2018-03-01 CRAN (R 3.5.0)                    
#>  rprojroot    1.3-2      2018-01-03 CRAN (R 3.5.0)                    
#>  scales       0.5.0      2017-08-24 CRAN (R 3.5.0)                    
#>  stats      * 3.5.0      2018-04-23 local                             
#>  stringi      1.1.7      2018-03-12 CRAN (R 3.5.0)                    
#>  stringr      1.3.0      2018-02-19 CRAN (R 3.5.0)                    
#>  tibble       1.4.2      2018-01-22 CRAN (R 3.5.0)                    
#>  tools        3.5.0      2018-04-23 local                             
#>  utils      * 3.5.0      2018-04-23 local                             
#>  withr        2.1.2      2018-03-15 CRAN (R 3.5.0)                    
#>  xml2         1.2.0      2018-01-24 CRAN (R 3.5.0)                    
#>  yaml         2.1.19     2018-05-01 CRAN (R 3.5.0)

[hadley]

@lionel- can you take a look at this one too please?

[lionel-]

@hadley What are the semantics for layers with data? should we have the ggplot data as an intermediate mask, so that class (from the tibble passed to ggplot()) would be available in the layer evaluation in all circumstances?

[hadley]

Hmmm, I think that's right — although to preserve the existing semantics that environment would binding every existing variable to NULL

[lionel-]

@mundl Are you sure this used to work? I get the same results with CRAN ggplot2.

@hadley We want the same evaluation semantics as in the last version right?

It doesn't seem merging the layer data frames would be much of an improvement since it only makes sense in the particular case where they have the same number of rows.

[lionel-]

Ah I get it now, this is about ignoring facet specifications that refer to variables that do not exist in the current layer.

I think we can pull this off by defining active bindings for all variables in the ggplot() layer that are missing in the current layer. These active bindings would send a condition that'd be picked up by the facet evaluation, signalling that facetting should be skipped for this layer.

Only downside: this prevents referring to variables from the environment. But you'll still be able to unquote those.

[hadley]

I think if they’re bound to NULL that will achieve the same effect. But if this isn’t a regression, there’s no need to work on now.

[lionel-]

Simple variables already work, this is about giving complex expressions the same semantics.

[lionel-]

When we fix this it'd be a good time to refactor the facetting code so the variables are evaluated only once. In the example above I see 5 evaluations of the vars() expression.