---
title: Japan population male female ratio by age and by prefecture
author: Mitsuo Shiota
date: '2025-03-26'
categories:
- economics
- R
knitr:
opts_chunk:
out.width: '90%'
---
```{r}
#| label: setup
#| message: false
library(tidyverse)
library(ggrepel)
library(sf)
theme_set(theme_light())
```
## Get data from e-stat
```{r}
#| label: get_data
estat_api_key <- Sys.getenv("ESTAT_API_KEY")
meta_info <- estatapi::estat_getMetaInfo(appId = estat_api_key,
lang = "J",
statsDataId = "0003448237")
dat_2023 <- estatapi::estat_getStatsData(appId = estat_api_key,
lang = "J",
statsDataId = "0003448237",
cdCat01 = c("001", "002"), # 男、女
cdCat03 = "001", # 総人口
cdTime = "1801" # 2023年10月1日現在
)
```
## Draw line charts
```{r}
#| label: line_charts
#| fig-cap: 男／女比率（全国＝1で標準化）2023年10月1日現在
#| fig-align: center
#| fig-asp: 1.5
dat_2023g <- dat_2023 |>
select(sex = "男女別", age = "年齢5歳階級", area_code, prefecture = "全国・都道府県", value) |>
mutate(
area_code = str_sub(area_code, 1L, 2L) |> as.numeric(),
region = case_when(
between(area_code, 1, 7) ~ "北海道・東北",
between(area_code, 8, 14) ~ "関東",
between(area_code, 15, 20) ~ "甲信越",
between(area_code, 21, 24) ~ "東海",
between(area_code, 25, 30) ~ "関西",
between(area_code, 31, 35) ~ "中国",
between(area_code, 36, 39) ~ "四国",
between(area_code, 40, 47) ~ "九州・沖縄",
.default = "全国"
)
) |>
pivot_wider(id_cols = c(age, area_code, prefecture, region),
names_from = sex, values_from = value) |>
rename(male = "男", female = "女") |>
mutate(ratio = male / female) |>
mutate(ratio_std = ratio / ratio[1], .by = age) |>
filter(region != "全国", age != "総数") |>
mutate(
age = fct_reorder(age, parse_number(age)),
region = fct_reorder(region, area_code)
)
dat_2023g |>
ggplot(aes(age, ratio_std)) +
geom_hline(yintercept = 1, color = "gray30", lty = 3) +
geom_line(aes(color = prefecture, group = prefecture), show.legend = FALSE) +
geom_text_repel(aes(label = prefecture), data = dat_2023g |> filter(age == "25～29歳"), size = 2, max.overlaps = 20) +
facet_wrap(vars(region), ncol = 2) +
labs(x = NULL, y = "男／女比率（全国＝1で標準化）") +
theme(axis.text.x = element_text(angle = 90))
```
## Choropleth map
```{r}
#| label: choropleth_map
#| fig-cap: 男／女比率（全国＝1で標準化）
#| fig-align: center
#| fig-asp: 1
prefectures_map <-
read_sf(system.file("shapes/jpn.shp", package = "NipponMap"))
# inset Okinawa
prefectures_map$geometry[[47]] <- prefectures_map$geometry[[47]] + c(7, 14)
lines <- cbind(c(132, 135, 137, 137), c(38, 38, 40, 43)) |>
st_linestring()
prefectures_map |>
mutate(jiscode = as.numeric(jiscode)) |>
left_join(dat_2023g,
by = join_by(jiscode == area_code)) |>
ggplot() +
geom_sf(aes(fill = ratio_std), color = "white") +
geom_sf(data = lines, color = "gray80") +
scale_fill_gradient2(low = "#559999", mid = "grey90", high = "#BB650B",
midpoint = 1) +
facet_wrap(vars(age), ncol = 4) +
labs(fill = "男／女比率（全国＝1で標準化）") +
coord_sf(xlim = c(130, 149), ylim = c(31, 45)) +
theme_void(base_size = 16) +
theme(legend.position = "bottom",
legend.title = element_text(size = 10),
legend.text = element_text(size = 6),
strip.text = element_text(size = 10))
```
