Require packages and summary population data

Taiwan population data source is from the following website: https://zh.wikipedia.org/wiki/臺灣行政區人口列表

require("dplyr")
require("stringr")
require("data.table")
require("ggplot2")
require("maptools")
require("knitr")
require("kableExtra")

df <- fread("D:\\tw_population.csv", h = T)
df <- data.frame(df)
kable(df)

縣市	類別	人口
新北市	直轄市	3985698
臺中市	直轄市	2783298
高雄市	直轄市	2776791
臺北市	直轄市	2683202
桃園市	直轄市	2181470
臺南市	直轄市	1886387
彰化縣	縣	1282934
屏東縣	縣	830697
雲林縣	縣	691021
苗栗縣	縣	554652
新竹縣	縣	551447
嘉義縣	縣	511797
南投縣	縣	501757
宜蘭縣	縣	456756
新竹市	市	440409
基隆市	市	371669
花蓮縣	縣	329462
嘉義市	市	269364
臺東縣	縣	219686
金門縣	縣	137042
澎湖縣	縣	103956
連江縣	縣	12823

summary(df)

##      縣市               類別                人口        
##  Length:22          Length:22          Min.   :  12823  
##  Class :character   Class :character   1st Qu.: 340014  
##  Mode  :character   Mode  :character   Median : 531622  
##                                        Mean   :1071014  
##                                        3rd Qu.:1735524  
##                                        Max.   :3985698

Popution visualization using a bar plot

The following steps aims to make a better plot. 1. transform population unit to 10,000 units 2. flip the coordinate 3. sort the cities by its population

colnames(df) <- c("city","category","population")
ggplot(df, aes(x = reorder(city, population), y = population/10000, fill = category)) + 
  geom_bar(stat="identity") +
  coord_flip() + 
  labs(title = "台灣縣市人口分布圖", x = "縣市", y = "人口數(萬)")

Read and plot Taiwan map data

Taiwan population data source is from the following website: http://www.gadm.org/country (shapefile form)

Read shape file
Tranform shp to dataframe
Plot Taiwan Map by ggplot2

taiwan_shp <- readShapeSpatial("D:\\TWN_adm_shp\\TWN_adm2.shp")

## Warning: use rgdal::readOGR or sf::st_read

## Warning: use rgdal::readOGR or sf::st_read

taiwan_map <- fortify(taiwan_shp)

## Regions defined for each Polygons

ggplot(taiwan_map, aes(x = long, y = lat, group=group)) +
  geom_path() + 
  coord_map()

Translate City Names from English to Chinese

#Sys.setlocale('LC_ALL','C')
print(as.character(taiwan_shp$NAME_2))

##  [1] "Kinmen"                    "Lienkiang (Matsu Islands)"
##  [3] "Hsinchu City"              "Kaohsiung"                
##  [5] "New Taipei City"           "Taichung"                 
##  [7] "Tainan"                    "Taipei"                   
##  [9] "Taoyuan"                   "Changhua"                 
## [11] "Chiayi City"               "Chiayi County"            
## [13] "Hsinchu County"            "Hualien"                  
## [15] "Keelung"                   "Miaoli"                   
## [17] "Nantou"                    "Penghu"                   
## [19] "Pingtung"                  "Taitung"                  
## [21] "Yilan"                     "Yulin"

chinese_name <- c("金門縣", "連江縣", "新竹市", "高雄市",
                  "新北市", "臺中市", "臺南市", "臺北市",
                  "桃園市", "彰化縣", "嘉義市", "嘉義縣",
                  "新竹縣", "花蓮縣", "基隆市", "苗栗縣",
                  "南投縣", "澎湖縣", "屏東縣", "臺東縣",
                  "宜蘭縣", "雲林縣")
print(chinese_name)

##  [1] "金門縣" "連江縣" "新竹市" "高雄市" "新北市" "臺中市" "臺南市"
##  [8] "臺北市" "桃園市" "彰化縣" "嘉義市" "嘉義縣" "新竹縣" "花蓮縣"
## [15] "基隆市" "苗栗縣" "南投縣" "澎湖縣" "屏東縣" "臺東縣" "宜蘭縣"
## [22] "雲林縣"

Combine these two Dataset for plotting

create a new data of cities
map population from “df” to “taiwan_shp”
merge two these two datasets, “mydata” and “taiwan_shp”

mydata <- data.frame(NAME_1=taiwan_shp$NAME_2,
                     NAME_2=chinese_name,
                     id=taiwan_shp$ID_2)
mydata$population <- 0
for(i in 1:nrow(mydata)){
  mydata$population[i] <- df$population[which(df$city == mydata$NAME_2[i])]
}
taiwan_map$id <- as.character(as.integer(taiwan_map$id)+1)
final.plot<-merge(taiwan_map,mydata,by="id",all.x=T)
head(final.plot)

##   id     long      lat order  hole piece group NAME_1 NAME_2 population
## 1  1 118.4654 24.44180     1 FALSE     1   0.1 Kinmen 金門縣     137042
## 2  1 118.4647 24.44181     2 FALSE     1   0.1 Kinmen 金門縣     137042
## 3  1 118.4646 24.44181     3 FALSE     1   0.1 Kinmen 金門縣     137042
## 4  1 118.4646 24.44153     4 FALSE     1   0.1 Kinmen 金門縣     137042
## 5  1 118.4640 24.44153     5 FALSE     1   0.1 Kinmen 金門縣     137042
## 6  1 118.4640 24.44125     6 FALSE     1   0.1 Kinmen 金門縣     137042

Plotting Taiwan Population Distribution Map

library(RColorBrewer) #配色用brewer.pal( 9 , "Reds" )
library(mapproj)
twcmap <- ggplot() +
  geom_polygon(data = final.plot, 
               aes(x = long, y = lat, group = group, 
                   fill = population/10000), 
               color = "black", size = 0.25) + 
  
  coord_map()+#維持地圖比例
  scale_fill_gradientn(colours = brewer.pal(9,"Reds"), name = "人口(萬)")+
  #theme_void()+
  labs(title="台灣縣市人口分佈圖", x ="經度", y = "緯度")

twcmap

Add City Locations Points with their names

search locations by google
add points and labels to the plotting
Another way to label

require(ggmap)
require(proto)
city <- geocode(str_c(as.character(mydata$NAME_2), ", 台灣"),
                source = "google")
city$name <- as.character(mydata$NAME_2)

city <- filter(city, lon !="")

twcmap + geom_point(aes(x=lon,y=lat),
                    color="forestgreen", size = 1.5, data=city) +
  geom_text(aes(x=lon,y=lat,label=name), data=city)

require(ggrepel)
twcmap + geom_label_repel(aes(x=lon,y=lat, label = name, alpha = 1),
                          data = city)

How to plot a Taiwan Map colored with Population by ggplot2

Ozu Shi

2018/01/13