R可视化：动态热力地图

# 设置为国内镜像, 方便快速安装模块
options("repos" = c(CRAN = "https://mirrors.tuna.tsinghua.edu.cn/CRAN/"))

wkdir <- '/home/user/R_workdir/TidyTuesday/2018/2018-04-02_US_Tuition_Costs/src-e'
setwd(wkdir)

library(ggtext)
library(tidyverse)
library(gganimate)
library(transformr)
library(fiftystater)
# 需要能访问 Google, 也可以注释掉下面这行, 影响不大
sysfonts::font_add_google("Gochi Hand", "gochi")

df_input <- readxl::read_excel("../data/us_avg_tuition.xlsx")

# 简要查看数据内容
glimpse(df_input)

## Rows: 50
## Columns: 13
## $ State     <chr> "Alabama", "Alaska", "Arizona", "Arkansas", "Caalbert", "C…
## $ `2004-05` <dbl> 5682.838, 4328.281, 5138.495, 5772.302, 5285.921, 4703.777, …
## $ `2005-06` <dbl> 5840.550, 4632.623, 5415.516, 6082.379, 5527.881, 5406.967, …
## $ `2006-07` <dbl> 5753.496, 4918.501, 5481.419, 6231.977, 5334.826, 5596.348, …
## $ `2007-08` <dbl> 6008.169, 5069.822, 5681.638, 6414.900, 5672.472, 6227.002, …
## $ `2008-09` <dbl> 6475.092, 5075.482, 6058.464, 6416.503, 5897.888, 6284.137, …
## $ `2009-10` <dbl> 7188.954, 5454.607, 7263.204, 6627.092, 7258.771, 6948.473, …
## $ `2010-11` <dbl> 8071.134, 5759.153, 8839.605, 6900.912, 8193.739, 7748.201, …
## $ `2011-12` <dbl> 8451.902, 5762.421, 9966.716, 7028.991, 9436.426, 8315.632, …
## $ `2012-13` <dbl> 9098.069, 6026.143, 10133.503, 7286.580, 9360.574, 8792.856,…
## $ `2013-14` <dbl> 9358.929, 6012.445, 10296.200, 7408.495, 9274.193, 9292.954,…
## $ `2014-15` <dbl> 9496.084, 6148.808, 10413.844, 7606.410, 9186.824, 9298.599,…
## $ `2015-16` <dbl> 9751.101, 6571.340, 10646.278, 7867.297, 9269.844, 9748.188,…
# 检查数据的列名
colnames(df_input)
##  [1] "State"   "2004-05" "2005-06" "2006-07" "2007-08" "2008-09" "2009-10"
##  [8] "2010-11" "2011-12" "2012-13" "2013-14" "2014-15" "2015-16"

# 调整年份为 YYYY 格式, 方便绘图展示
df_tidy <- df_input %>%
  dplyr::rename(
    "2004" = "2004-05",
    "2005" = "2005-06",
    "2006" = "2006-07",
    "2007" = "2007-08",
    "2008" = "2008-09",
    "2009" = "2009-10",
    "2010" = "2010-11",
    "2011" = "2011-12",
    "2012" = "2012-13",
    "2013" = "2013-14",
    "2014" = "2014-15",
    "2015" = "2015-16")

# 第一张图片数据预处理
df_tidy.1 <- df_tidy %>%
  # 建议使用 dplyr::mutate 形式调用函数, 有可能与 plyr 中的函数冲突 (因为我自己就报错了...)
  dplyr::mutate(id = tolower(State)) %>%
  dplyr::mutate(cagr = ((`2015`/`2004`)**(1/11)-1)*100) %>%
  select(id, cagr)

# 第二张图片数据预处理
df_tidy.2 <- df_tidy %>%
  dplyr::mutate(id = tolower(State)) %>%
  select(-State) %>%
  gather(key = year, value = cost, -id)


# 简要查看数据内容
glimpse(df_tidy.1)
## Rows: 50
## Columns: 2
## $ id   <chr> "alabama", "alaska", "arizona", "arkansas", "california", "colora…
## $ cagr <dbl> 5.0309094, 3.8688482, 6.8464675, 2.8549355, 5.2391738, 6.8490463,…
glimpse(df_tidy.2)
## Rows: 600
## Columns: 3
## $ id   <chr> "alabama", "alberta", "arizona", "arkansas", "california", "colora…
## $ year <chr> "2004", "2004", "2004", "2004", "2004", "2004", "2004", "2004", "…
## $ cost <dbl> 5682.838, 4328.281, 5138.495, 5772.302, 5285.921, 4703.777, 7983.…

# PS: 方便讲解, 我这里进行了拆解, 具体使用时可以组合在一起
gg <- ggplot(df_tidy.1, aes(map_id = id))
# geom_map() 绘制地图
gg <- gg + geom_map(aes(fill = cagr), color = "azure4", map = fifty_states)
# expand_limits() 控制坐标轴的范围
gg <- gg + expand_limits(x = fifty_states$long, y = fifty_states$lat)
# coord_map() 将地图变成球状展示
gg <- gg + coord_map()
# scale_x_continuous() 对连续变量设置坐标轴显示范围
gg <- gg + scale_x_continuous(breaks = NULL)
# scale_y_continuous() 对连续变量设置坐标轴显示范围
gg <- gg + scale_y_continuous(breaks = NULL)
# geom_text() 添加文本信息
gg <- gg + geom_text(x = -68.5, y = 36.5, label = "CAGR (%)", size = 4) 
gg <- gg + ggthemes::theme_fivethirtyeight()
# scale_fill_continuous() 针对连续型数据, 自定义图像的色彩梯度
gg <- gg +scale_fill_continuous(name = "", low = "#00FF00", high = "#FF0000")
# labs() 对图形添加注释和标签（包含标题、子标题、坐标轴和引用等注释)
gg <- gg + labs(title = "从2004年到2016年美国各种学费的年均变化",
                x = NULL,
                y = NULL,
                caption = "资料来源: <i>onlinembapage.com</i> · graph by 萤火之森")
# theme_minimal() 去坐标轴边框的最小化主题
gg <- gg + theme_minimal()
# theme() 实现对非数据元素的调整, 对结果进行进一步渲染, 使之更加美观
gg <- gg + theme(
  # plot.margin 调整图像边距, 上-右-下-左
  plot.margin = grid::unit(c(9, 16, 9, 16), "mm"),
  # aspect.ratio 固定图像的纵横比
  aspect.ratio = 9/16,
  # panel.background 面板背景 数据下面
  panel.background = element_blank(),
  # plot.title 主标题
  plot.title = element_text(hjust = 0.5, size = 32),
  # plot.caption 说明文字
  plot.caption = element_markdown(colour = "dodgerblue4", size = 12),
  # legend.text 设置图例文本格式
  legend.text = element_text(size = 12),
  # legend.title 设置图例标题
  legend.title = element_text(size = 12),
  # legend.position 设置图例位置, 这里用坐标来指定图例的位置
  legend.position = c('.88', '.2'),
  # legend.key.size 调整图例的大小
  legend.key.size = unit(.75, 'cm'),
  # plot.background 图片背景
  plot.background = element_rect(fill = "white"))

gg

filename = '20180402-E-01'
ggsave(filename = paste0(filename, ".pdf"), width = 12.2, height = 7.6, device = cairo_pdf)
ggsave(filename = paste0(filename, ".png"), width = 12.2, height = 7.6, dpi = 100, device = "png")

# PS: 方便讲解, 我这里进行了拆解, 具体使用时可以组合在一起
gg <- ggplot(df_tidy.2, aes(frame = year, map_id = id, group = interaction(year, id)))
# geom_map() 绘制地图
gg <- gg + geom_map(aes(fill = cost), color = "azure4", map = fifty_states)
# expand_limits() 控制坐标轴的范围
gg <- gg + expand_limits(x = fifty_states$long, y = fifty_states$lat)
# coord_map() 将地图变成球状展示
gg <- gg + coord_map()
# scale_x_continuous() 对连续变量设置坐标轴显示范围
gg <- gg + scale_x_continuous(breaks = NULL)
# scale_y_continuous() 对连续变量设置坐标轴显示范围
gg <- gg + scale_y_continuous(breaks = NULL, labels = function(x) format(x, big.mark = ".", scientific = FALSE))
# transition_states() 在动画中数据的几个不同阶段之间的转换
gg <- gg + transition_states(year, transition_length = 1, state_length = 4)
# guides() 设置图例信息
gg <- gg + guides(fill = guide_legend(title = "学费及相关的费用"))
# scale_fill_gradientn() 将颜色比例转换为概率转换颜色分布
gg <- gg + scale_fill_gradient(low = "#00FF7F", high = "#DC143C")
gg <- gg + ggthemes::theme_fivethirtyeight() 
# labs() 对图形添加注释和标签（包含标题、子标题、坐标轴和引用等注释)
gg <- gg + labs(title = "美国各州的平均学费：{closest_state} 年",
                subtitle = '',
                x = NULL,
                y = NULL,
                caption = "资料来源: onlinembapage.com · graph by 萤火之森")
# theme_minimal() 去坐标轴边框的最小化主题
gg <- gg + theme_minimal()
# theme() 实现对非数据元素的调整, 对结果进行进一步渲染, 使之更加美观
gg <- gg + theme(
  # panel.grid.major 主网格线, 这一步表示删除主要网格线
  panel.grid.major = element_line("grey", size = 0.2),
  # panel.grid.minor 次网格线, 这一步表示删除次要网格线
  panel.grid.minor = element_blank(),
  # plot.margin 调整图像边距, 上-右-下-左
  plot.margin = grid::unit(c(9, 16, 9, 16), "mm"),
  # aspect.ratio 固定图像的纵横比
  aspect.ratio = 9/16,
  # panel.background 面板背景 数据下面
  panel.background = element_blank(),
  # plot.title 主标题
  plot.title = element_text(family = 'gochi',hjust = 0.5, size = 40),
  # plot.subtitle 次要标题
  plot.subtitle = element_markdown(size = 20),
  # plot.caption 说明文字
  plot.caption = element_markdown(colour = "red", size = 24),
  # legend.text 设置图例文本格式
  legend.text = element_text(family = 'gochi',size = 20),
  # legend.title 设置图例标题
  legend.title = element_text(family = 'gochi',size = 20),
  # legend.position 设置图例位置, 这里用坐标来指定图例的位置
  legend.position = c('.79', '.92'),
  # legend.key.size 调整图例的大小
  legend.key.size = unit(.75, 'cm'),
  # legend.box.background 设置图例背景
  legend.box.background = element_blank(),
  # plot.background 图片背景
  plot.background = element_blank())

# 通过调整 duration 参数设置动画长度
animate(gg, renderer = gifski_renderer(), width = 1580, height = 1000, duration = length(unique(df_tidy.2$year)))

filename = '20180402-E-02'
anim_save(filename = paste0(filename, ".gif"))