RからeStatのAPIを使ってみる

library(pacman)
p_load(keyring, httr, listviewer, rlist, pipeR, dplyr, stringr, ggplot2, stringi,
       readr, ggthemes)
keyring::key_set("e-stat")

## 統計表を検索してお目当ての表を特定
response <- GET(
  url = "https://api.e-stat.go.jp/rest/2.1/app/json/getStatsList",
  query = list(
    appId = keyring::key_get("e-stat"),
    searchWord = "障害種 AND 状況別 AND 高等部"
  )
)

res_content <- content(response)
jsonedit(res_content)

id_table <- res_content$GET_STATS_LIST$DATALIST_INF$TABLE_INF %>>%
  list.select(
    id = `@id`,
    year = STATISTICS_NAME_SPEC$TABULATION_SUB_CATEGORY1,
    table_name = TITLE_SPEC$TABLE_NAME) %>>%
  list.stack()
id_list <- pull(id_table, id)

res <- tibble()

for (id in id_list){
  dat_year <- GET(
    url = "https://api.e-stat.go.jp/rest/2.1/app/getSimpleStatsData",
    query = list(
      appId = keyring::key_get("e-stat"),
      statsDataId = id
    )
  )

  content_year <- content(dat_year)

  dat_year <-read.csv(text = sub('(?s).*"VALUE"\n', "", content_year, perl = TRUE))
  dat <- dat_year %>%
    rename("disability" = contains("障害種"),
           "sex" = contains("性別"),
           "path" = contains("状況"),
           "year"=contains("年次")) %>%
    mutate(path=as.character(path)) %>%
    filter(disability=="知的障害") %>%
    filter(!str_detect(sex, "男" )) %>%
    filter(!str_detect(sex, "女" )) %>%
    filter(!str_detect(path,"再掲")) %>%
    filter(!str_detect(path, "％")) %>%
    filter(!str_detect(path, "除く")) %>%
    filter(path !="計") %>%
    mutate_if(is.factor, as.character) %>%
    mutate(value=as.character(value))

  if(is.null(dat$year)){dat <- mutate(dat, year="2010年")}

  dat <- dat %>%
    select(year ,path, value)

  res<- bind_rows(res, dat)
}  

res <- arrange(res, year) %>%
  mutate(value=as.integer(value))

write_csv(res,"path_hs.csv")

## ここから積み上げグラフを作る
dat <- read_csv("path_hs.csv", col_types = "cci")
lev <- c("大学等進学者（A）",
        "専修学校（専門課程）進学者（B）",
        "専修学校（一般課程）等入学者（C）",
        "公共職業能力開発施設等入学者（D）",
        "就職者",
        "左記以外の者",
        "不詳・死亡の者")
dat_mo <- dat %>%
  mutate(year=as.factor(year),path=factor(path,levels = lev))

p <- ggplot(dat_mo, aes(x=year, y=value, fill=path)) +
  geom_bar(stat = "identity") +
  labs(x="年度", y="人数") +
  theme_gdocs(base_family="HiraKakuPro-W3") +
  scale_fill_gdocs(name="卒業後の状況") +
  ggtitle("特別支援学校高等部卒業後の状況（知的障害）")+
  labs(caption="出典「文部科学省　学校基本調査」") +
  theme(plot.caption=element_text(size=8, hjust=0,))
print(p)
ggsave("images/graduate_hs.png",p, width=10, height=5, dpi=300)

## 割合のグラフ
dat_per <- dat_mo %>%
  group_by(year) %>%
  mutate(total = sum(value)) %>%
  mutate(percent=round(100*(value/total),3))
br <- seq(0,100, 10)
p1 <- ggplot(dat_per, aes(x=year, y=percent, fill=path)) +
  geom_bar(stat = "identity") +
  scale_y_continuous(breaks=br) +
  labs(x="年度", y="割合（％）") +
  theme_gdocs(base_family="HiraKakuPro-W3") +
  scale_fill_gdocs(name="卒業後の状況") +
  ggtitle("特別支援学校高等部卒業後の状況（知的障害）") +
  labs(caption="出典「文部科学省　学校基本調査」") +
  theme(plot.caption=element_text(size=8, hjust=0,))
print(p1)
ggsave("images/graduate_hs_per.png",p1, width=10, height=5, dpi=300)