facing difficulty for reproducing maps in R

Question

I am struggling for several months to reproduce a map that is saved in the folder (codes_that_following) but did not find a solution to make it. I am also sharing the original codes that I want to reproduce for my study area, find the link below;

https://www.dropbox.com/sh/yj8ie1qj2w0tutu/AACigc0Qk0wNoUGVHxF1Ly0oa?dl=0

when I run the codes,

ggplot(fips_us) + 
  geom_sf(aes(fill = carbon, geometry = geometry), color = NA) +
  geom_sf(data = us_outline, aes(geometry = geometry), 
           fill = NA, size = 0.25) + 
    scale_fill_scico(palette = "bilbao") +
    facet_grid(vars(month))

it gives me this error

Error: cannot allocate vector of size 0 Kb Error during wrapup: memory exhausted (limit reached?) 
Error: no more error handlers available (recursive errors?); invoking 'abort' restart

Answer 1

Two problems with your code.

1. You are grouping by lat long, and your are not using points to describe the underlying data, but shapes (states), so when you right_join with the sf object it creates a new geometry for each lat long. And that is too much.

Solution: grup by state NAME_1

2. The geometries themselves are too much detailed. Specially if you showing them from so "far away", it´s just useless information and makes is harder for ggplot2 to compute.

Solution: use rmapshaper::ms_simplify to simplify all the geometries.

Bellow the code and output:

library(tidyverse)
library(sf)
library(rmapshaper)

# Read in drought data
drought_fips <- readxl::read_xlsx("willdfire_climate_data.xlsx")
chi <- sf::read_sf("gadm36_CHN_1.shp")

# simplify
chi <- rmapshaper::ms_simplify(chi, keep = 0.001)

# Calculate mean DCSI by year and month
month_fips <- drought_fips %>% 
  group_by(year, month, NAME_1) %>% 
  summarise(carbon = mean(fire.carbon))

# Combine with shapefile
fips_chi <- chi %>% 
  right_join(month_fips)

# select a few years and months for example
# stackoverflow has a 2mb limit for images, and your image with all the years and months has to be very large to be properly readable. So in your code use fips_chi and save it with large dimensions in ggsave.
fips_chi2 <- fips_chi %>%  
  filter(year %in% c(2015:2019),
         month %in% c(1:6))
  
# plot
g <- ggplot() +
  geom_sf(data = chi,
          aes(geometry = geometry),
          color = "black",
          fill = "white",
          stroke = 0.1) +   
  geom_sf(data = fips_chi2,
          aes(fill = carbon, 
              geometry = geometry),
          stroke = 0.1) +
  scale_fill_viridis_c(labels = scales::number_format(scale = 1e-3),
                       name = "Carbon\nthousands") +
  facet_grid(year ~ month) +
  theme_bw() +
  theme(panel.background = element_rect(fill = 'white', colour = 'white'),
        axis.line = element_blank(), 
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        axis.title = element_blank(),
        panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        panel.spacing = unit(0, "lines")) +
  guides(fill = guide_colourbar(barwidth = 1, barheight = 12))


g
# save plot
ggsave(filename = "Carbon9.png", 
       height = 8, 
       width = 12)

I belive you be able to further customize this example to your needs.