geom_spatraster dropping levels when plotting

[mooibroekd]

Hi,

I have a spatraster that looks like this:

class       : SpatRaster 
dimensions  : 320, 1095, 1  (nrow, ncol, nlyr)
resolution  : 0.2, 0.2  (x, y)
extent      : -125, 94, 23, 87  (xmin, xmax, ymin, ymax)
coord. ref. : lon/lat WGS 84 (EPSG:4326) 
source(s)   : memory
categories  : label 
name        : lyr.1 
min value   :  0-10 
max value   : 30-60 

However, there are more levels defined in the SpatRaster:

> terra::levels(r1)[[1]]
  ID  label
1  0   0-10
2  1  10-30
3  2  30-60
4  3  60-90
5  4 90-120
> terra::cats(r1)[[1]]
  ID  label
1  0   0-10
2  1  10-30
3  2  30-60
4  3  60-90
5  4 90-120
> terra::is.factor(r1)
[1] TRUE

I am using geom_spatraster to plot this and apply a ggplot2::scale_fill_manual call to set the correct colors. The values provided are:

> colorVal
     0-10     10-30     30-60     60-90    90-120 
"#2892c7" "#a0c79b" "#fafa64" "#fa8d34" "#e81014

While the legend labels exists for the last two bins, the colors in the legend will not show. Only the first three bins, that are present in the data, are shown.

terra::plot has the same behaviour, only plotting what is available. It also has a all_levels variable, which set to TRUE plots all levels in the SpatRaster. Hence, I suspect geom_spatraster is not using the levels, only the data.

My question would be, how can I still show all the legend items, even if levels are not present in the data?

[dieghernan]

See https://stackoverflow.com/a/45765225/7877917 use show.legend = TRUE in the geom and drop = FALSE in the scale:

library(tidyterra)
#> 
#> Adjuntando el paquete: 'tidyterra'
#> The following object is masked from 'package:stats':
#> 
#>     filter
library(tidyverse)
library(terra)
#> terra 1.8.42
#> 
#> Adjuntando el paquete: 'terra'
#> The following object is masked from 'package:tidyr':
#> 
#>     extract
library(ggplot2)


r_init <- rast(system.file("ex/elev.tif", package = "terra"))

r1 <- r_init %>%
  mutate(
    cats = scales::rescale(elevation, to = c(0, 60)),
    lyr.1 = cut(cats,
      breaks = c(0, 10, 30, 60, 90, 120),
      labels = c("0-10", "10-30", "30-60", "60-90", "90-120")
    )
  ) %>%
  select(lyr.1)

r1
#> class       : SpatRaster 
#> dimensions  : 90, 95, 1  (nrow, ncol, nlyr)
#> resolution  : 0.008333333, 0.008333333  (x, y)
#> extent      : 5.741667, 6.533333, 49.44167, 50.19167  (xmin, xmax, ymin, ymax)
#> coord. ref. : lon/lat WGS 84 (EPSG:4326) 
#> source(s)   : memory
#> categories  : label 
#> name        : lyr.1 
#> min value   :  0-10 
#> max value   : 30-60

terra::levels(r1)[[1]]
#>   value  label
#> 1     1   0-10
#> 2     2  10-30
#> 3     3  30-60
#> 4     4  60-90
#> 5     5 90-120
terra::cats(r1)[[1]]
#>   value  label
#> 1     1   0-10
#> 2     2  10-30
#> 3     3  30-60
#> 4     4  60-90
#> 5     5 90-120
terra::is.factor(r1)
#> [1] TRUE

colorVal <- c("#2892c7", "#a0c79b", "#fafa64", "#fa8d34", "#e81014")
names(colorVal) <- levels(r1$lyr.1)[[1]]$label

colorVal
#>      0-10     10-30     30-60     60-90    90-120 
#> "#2892c7" "#a0c79b" "#fafa64" "#fa8d34" "#e81014"

ggplot() +
  geom_spatraster(data = r1, show.legend = TRUE) +
  scale_fill_manual(values = colorVal, drop = FALSE, na.translate = FALSE)

^{Created on 2025-04-24 with reprex v2.1.1}

[mooibroekd]

Thanks, I was aware of the drop = FALSE in the scale_fill_manual call. The show.legend was something I missed and I suspect that was the culprit. I replaced the terra::classify with the tidyterra cut option, which gives what I want without changing the levels in the classified raster again.

+1 for using my desired color scheme.