`geom_spatraster()` gives : Error in `ggplot2::scale_fill_manual()` with raster

[beausoleilmo]

I was trying the package and wasn't able to make geom_spatraster() work

From the tif data here :
https://www.donneesquebec.ca/recherche/fr/dataset/utilisation-du-territoire

lulc_raw = rast(x = "utilisation_territoire_2020/utilisation_territoire_2020.tif")
gg_map_gbif = ggplot() + 
geom_spatraster(data = lulc_raw) 
<SpatRaster> resampled to 500621 cells.

 gg_map_gbif
Error in `ggplot2::scale_fill_manual()`:
! Continuous values supplied to discrete scale.
ℹ Example values: NA, NA, NA, NA, and NA
Run `rlang::last_trace()` to see where the error occurred.

Is the error (Error in ggplot2::scale_fill_manual()) because the tif has a cats table?

[dieghernan]

Yup, there's something messy with the coltab, and I think is in the source of the tif. But if you transform it to factor (as.factor()) it seems to work. however your data is huge, couldn't do a proper test. See my testing script, restricting the area to Ile de Montreal:

# Restrict for reprex
library(terra)
library(tidyterra)
library(ggplot2)

lulc_raw = rast(x = "utilisation_territoire_2020/utilisation_territoire_2020.tif")


sample_area <- nominatimlite::geo_address_lookup_sf("3430571", "r", points_only = FALSE) %>% 
  sf::st_transform(pull_crs(lulc_raw))

sample_area

# Simple feature collection with 1 feature and 2 fields
# Geometry type: POLYGON
# Dimension:     XY
# Bounding box:  xmin: -428735.3 ymin: 172030.2 xmax: -387453.1 ymax: 203200.8
# Projected CRS: NAD83 / Quebec Lambert
# # A tibble: 1 × 3
#  query    address                                                       geometry
# * <chr>    <chr>                                                    <POLYGON [m]>
# 1 r3430571 Île de Montréal, Agglomération de Montréal,… ((-428735.3 174717.5, -4…

# Not working
crop_smp <- crop(lulc_raw, sample_area)
ggplot() + 
  geom_spatraster(data = crop_smp, na.rm = TRUE) +
  guides(fill = "none")

# <SpatRaster> resampled to 500610 cells.
# Error in `ggplot2::scale_fill_manual()`:
# ! Continuous values supplied to discrete scale.
# ℹ Example values: 55, 565, 55, 158, and 158
# Run `rlang::last_trace()` to see where the error occurred.

# But...
crop_smp2 <- as.factor(crop_smp)

ggplot() + 
  geom_spatraster(data = crop_smp2, na.rm = TRUE) +
  guides(fill = "none")

# <SpatRaster> resampled to 500610 cells.

[dieghernan]

Here you have different alternatives to plot rasters with color tables, hopefull it is useful https://dieghernan.github.io/tidyterra/articles/faqs.html#plotting-a-spatraster-with-color-tables