跨地理方面的ggplot背景图像 完整示例代码

编程问答 2022-05-16

问题描述

我想在左侧图表的背景中添加一张真实的欧洲地图：

enter image description here

原始图的代码取自 here（生成 this graph at "03/08/2020 ? European Power"），并且有一个略短的版本，其中包含我建议的解决方案，但在本文结尾处无法按预期工作。简而言之，我的建议是：

第一步：获取图片

img_path <- "https://upload.wikimedia.org/wikipedia/commons/thumb/6/66/Blank_map_of_Europe_cropped.svg/1000px-Blank_map_of_Europe_cropped.svg.png"
img <- readpnG(img_path)
g <- rasterGrob(img,interpolate = FALSE)

第 2 步：按照建议将图像添加到 ggplot-background here

annotation_custom(g,-Inf,Inf,Inf)

但它并没有完成这项工作。有人知道解决这个问题的方法吗？

编辑： 为清楚起见，一旦地图出现，我不想进一步调整网格以使每组条形与一个国家/地区的位置完全匹配。我真的只想用右边的地图替换左边的深色背景。

完整示例代码

pacman::p_load(tidyverse,tidytuesdayR,geofacet,grid,png)

## import image
img_path <- "https://upload.wikimedia.org/wikipedia/commons/thumb/6/66/Blank_map_of_Europe_cropped.svg/1000px-Blank_map_of_Europe_cropped.svg.png"
img <- readpnG(img_path)
g <- rasterGrob(img,interpolate = FALSE)

## Load in data
tuesdata <- tidytuesdayR::tt_load('2020-08-04')

## Match countries in data w/ grid,do some pivoting
energy_raw <- tuesdata$energy_types %>%
  mutate(country_name = case_when(country_name == "north Macedonia" ~ "N. Macedonia",country_name == "Bosnia & Herzegovina" ~ "Bosnia & H.",country == "UK" ~ "UK",country == "EL" ~ "Greece",TRUE ~ country_name)) %>%
  filter(country_name %in% europe_countries_grid2$name) %>%
  pivot_longer(cols = c(`2016`,`2017`,`2018`),names_to = "year") %>%
  pivot_wider(names_from = type,values_from = value) %>%
  select(-country)

## Work out percentages of the total and tidy the data
energy_plot <- left_join(energy_raw %>% filter(level == "Level 1") %>% select(-level) %>% janitor::remove_empty("cols"),energy_raw %>% filter(level == "Level 2") %>% select(-level) %>% janitor::remove_empty("cols")) %>% 
  janitor::clean_names() %>%
  rowwise() %>%
  mutate(tot = sum(conventional_thermal,nuclear,hydro,wind,solar,geothermal,other,pumped_hydro_power),across(where(is.numeric),~ 100 * . / tot)) %>%
  select(-tot) %>%
  ungroup() %>%
  pivot_longer(names_to = "energy",values_to = "value",c(conventional_thermal,pumped_hydro_power)) %>%
  rbind(expand.grid(country_name = setdiff(europe_countries_grid2$name,energy_raw$country_name),year = 2016:2018,energy = "No Data Available",value = 100)) %>%
  mutate(energy = case_when(energy == "conventional_thermal" ~ "Conventional Thermal",energy == "nuclear" ~ "Nuclear",energy == "No Data Available" ~ energy,TRUE ~ "Renewable"),energy = factor(energy,levels = c("Conventional Thermal","Nuclear","Renewable","No Data Available"))) %>%
  group_by(energy,year,country_name) %>%
  summarise(value = sum(value)) %>% mutate(country_name = str_trunc(country_name,11))

## Plot!

ggplot(energy_plot) +
  
  # background image
  annotation_custom(g,Inf) + 
  
  # The actual proper plotting stuff here isn't actually that complicated!
  aes(x = year,y = value,fill = energy) +
  geom_col(color = "#30332E") +
  coord_flip() +
  scale_y_reverse() +
  
  # facet_geo from the "geofacet" package - I truncate the names here just to be sure
  facet_geo( ~ country_name,grid = europe_countries_grid2 %>% mutate(name = str_trunc(name,11))) +
  
  # colours and themes
  scale_fill_manual(values = c("#EF476F","#FFD166","#06D6A0","#4B5446")) +
  theme_minimal() +
  theme(
    panel.grid = element_blank(),axis.text.x = element_blank(),plot.margin = unit(c(1,1,1.5,1.2),"cm"),legend.position = c(0.075,.125)
  )

解决方法

暂无找到可以解决该程序问题的有效方法，小编努力寻找整理中！

如果你已经找到好的解决方法，欢迎将解决方案带上本链接一起发送给小编。

小编邮箱:dio#foxmail.com (将#修改为@）

background-imagefacet-gridggplot2rr