blackdown_typhoon15_chiba1909.R

#############################################
# 東京電力街が提供する千葉県内の停電状況
#############################################
library(dplyr)
library(sf)
library(ggplot2)
library(rcartocolor)
library(ggrepel)
library(gridExtra)

if (file.exists(here::here("data-raw/tepco_blackout_20190910_2300_pref12.rds")) == FALSE) {
  library(rvest)
  site_url <- "http://teideninfo.tepco.co.jp/html"
  x <-
    glue::glue(site_url, "/12000000000.html") %>%
    read_html()
  x %>%
    html_nodes(css = "#main_container > p:nth-child(3)") %>%
    html_text(trim = TRUE)
  df_pref12_blackout <-
    x %>%
    html_nodes(css = "#main_container > table.bo_lv2 > tbody > tr") %>%
    purrr::map_dfr(
      ~ tibble::tibble(
        city_name = .x %>% html_nodes(css = "td.bo_lv2_occur > a") %>% html_text(),
        value = .x %>% html_nodes(css = "td.bo_lv2_occur_r") %>% html_text(),
        url = .x  %>% html_nodes(css = "td.bo_lv2_occur > a")%>% html_attr("href"))) %>%
    mutate(url = paste(site_url, url, sep = "/"))
  collect_city_blackout_table <- function(url) {
    x <-
      url %>%
      read_html()
    tibble::tibble(
      city_code = stringr::str_remove(basename(url), ".html"),
      city_name = x %>%
        html_nodes(css = "#area_left") %>%
        html_text(trim = TRUE) %>%
        stringr::str_split("-", simplify = TRUE) %>%
        purrr::pluck(3) %>%
        stringr::str_trim(),
      name = x %>%
        html_nodes(css = "#main_container > table.bo_lv3 > tbody > tr > td:nth-child(odd)") %>%
        html_text(trim = TRUE),
      value = x %>%
        html_nodes(css = "#main_container > table.bo_lv3 > tbody > tr > td:nth-child(even)") %>%
        html_text(trim = TRUE)
    ) %>%
      filter(name != "")
  }

  slw_collect_city_blackout_table <-
    purrr::slowly(~ collect_city_blackout_table(.x),
                  rate = purrr::rate_delay(pause = 3),
                  quiet = FALSE)
  df_pref12_blackout_town <-
    df_pref12_blackout %>%
    pull(url) %>%
    purrr::map_dfr(slw_collect_city_blackout_table)

  df_pref12_blackout_town <-
    df_pref12_blackout_town %>%
    mutate(under = stringr::str_detect(value, "未満")) %>%
    mutate(value = readr::parse_number(value)) %>%
    group_by(city_code, city_name, name) %>%
    mutate(value = sum(value, na.rm = TRUE)) %>%
    slice(1L) %>%
    ungroup()
  df_pref12_blackout_town %>%
    readr::write_rds(here::here("data-raw/tepco_blackout_20190910_2300_pref12.rds"))
} else {
  df_pref12_blackout_town <-
    readr::read_rds(here::here("data-raw/tepco_blackout_20190910_2300_pref12.rds"))
}

# 地図 ----------------------------------------------------------------------
estat_district <- function(path) {
  d <-
    sf::st_read(path,
                stringsAsFactors = FALSE,
                as_tibble = TRUE) %>%
    janitor::clean_names() %>%
    dplyr::select(-area, -perimeter, -ken, -ken_name, -dummy1, -area_max_f,
                  -n_ken, -n_city, -kbsum, -x_code, -y_code, -kcode1,
                  -h27k_axx_id, -h27k_axx,
                  -kigo_e, -kigo_d, -kigo_i, -hcode,
                  -keycode1, -keycode2)
  d <-
    d %>%
    dplyr::select(pref, pref_name, city, city_name, s_name, kihon1, jinko, setai) %>%
    sf::st_drop_geometry() %>%
    dplyr::group_by(city, kihon1) %>%
    dplyr::mutate(jinko = sum(jinko),
                  setai = sum(setai)) %>%
    dplyr::slice(1L) %>%
    dplyr::ungroup() %>%
    dplyr::left_join(d %>%
                       dplyr::group_by(city, kihon1) %>%
                       dplyr::summarise() %>%
                       dplyr::ungroup(),
                     by = c("city", "kihon1")) %>%
    sf::st_sf() %>%
    dplyr::mutate_if(is.character, stringi::stri_trans_nfkc)
  d %>%
    dplyr::mutate(s_name = if_else(stringr::str_detect(s_name, "丁目$"),
                                   stringr::str_remove(s_name, "([一二三四五六七八九十壱弐参拾百千万萬億兆〇]|[0-9])丁目$"),
                                   s_name),
                  s_name = stringr::str_remove_all(s_name, "[A-Z]$"))
}

sf_pref12 <-
  estat_district("~/Documents/resources/e-Stat/2015_国勢調査_小地域/A002005212015DDSWC12/h27ka12.shp") %>%
  filter(s_name != "水面", !is.na(s_name)) %>%
  group_by(city, city_name, s_name) %>%
  summarise(setai = sum(setai, na.rm = TRUE)) %>%
  ungroup() %>%
  select(city_name, s_name, setai) %>%
  left_join(df_pref12_blackout_town %>%
              mutate(city_name = recode(city_name,
                                        `袖ケ浦市` = "袖ヶ浦市")),
            by = c("city_name",
                                            "s_name" = "name")) %>%
  select(names(.)[!names(.) %in% attr(., "sf_column")])

p1 <-
  sf_pref12 %>%
  ggplot() +
  geom_sf(aes(fill = value), color = "gray", size = 0.08) +
  scale_fill_carto_c(palette = "OrYel", na.value = "white") +
  theme_void(base_family = "IPAexGothic", base_size = 12) +
  coord_sf(datum = NA) +
  guides(fill = guide_colorbar(title = "停電件数\n（「約」および「未満」を含む）",
                               title.position = "top",
                               title.vjust = 0.95)) +
  theme(legend.key.height = unit(3.0, "line"),
        plot.caption = element_text(size = 6)) +
  labs(title = "東京電力供給による千葉県内 地域別停電件数",
       subtitle = "2019年9月10日 22:59現在",
       caption = "データソース: http://teideninfo.tepco.co.jp/index-j.html\n千葉県行政区域: http://e-stat.go.jp/")

p2 <-
  sf_pref12 %>%
  mutate(city_name_s_name = paste(s_name, "\n(", city_name, ")")) %>%
  top_n(10, value) %>%
  ggplot(aes(forcats::fct_reorder(city_name_s_name, value), value, fill = value)) +
  scale_fill_carto_c(palette = "OrYel", na.value = "white") +
  geom_bar(stat = "identity") +
  geom_text(aes(label = value), nudge_y = -600, color = "white") +
  theme_bw(base_family = "IPAexGothic", base_size = 12) +
  coord_flip() +
  xlab(NULL) +
  ylab("停電件数") +
  guides(fill = FALSE) +
  theme(legend.key.height = unit(3.0, "line")) +
  labs(title = "東京電力供給による千葉県内 地域別停電件数 上位10地域",
       subtitle = "2019年9月10日 22:59現在",
       caption = "データソース: http://teideninfo.tepco.co.jp/index-j.html")

ggsave(filename = "tepco_blackout_20190910_2300_pref12_area.png",
       plot = gridExtra::grid.arrange(p2, p1, nrow = 1),
       width = 12, height = 5,
       dpi = "retina")

sf_pref12_city <-
  sf_pref12 %>%
  group_by(city_name) %>%
  summarise(value = sum(value, na.rm = TRUE)) %>%
  mutate(value = if_else(value == 0, NA_real_, value)) %>%
  mutate(value_c = case_when(
    value < 1000 ~ "〜1000件",
    between(value, 1001, 10000) ~ "1001〜10000件",
    value > 10001 ~ "10001件〜"
  ))

p3 <-
  sf_pref12_city %>%
  ggplot() +
  geom_sf(aes(fill = value_c), color = "gray", size = 0.08) +
  geom_sf_text(data = . %>% filter(!is.na(value)),
               aes(label = city_name), size = 1.2, family = "IPAexGothic") +
  scale_fill_manual(values = c("〜1000件" = rgb(250, 245, 0, maxColorValue = 255),
                               "1001〜10000件" = rgb(255, 153, 0, maxColorValue = 255),
                               "10001件〜" = rgb(255, 40, 0, maxColorValue = 255))) +
  theme_void(base_family = "IPAexGothic", base_size = 12) +
  coord_sf(datum = NA) +
  guides(fill = guide_legend(title = "停電件数\n（「約」および「未満」を含む）",
                               title.position = "top",
                               title.vjust = 0.95)) +
  theme(legend.key.height = unit(2.0, "line")) +
  labs(title = "東京電力供給による千葉県内 地域別停電件数",
       subtitle = "2019年9月10日 22:59現在",
       caption = "データソース: http://teideninfo.tepco.co.jp/index-j.html")

ggsave(filename = "tepco_blackout_20190910_2300_pref12_city.png",
       plot = p3,
       width = 7, height = 5,
       dpi = "retina")