R language draws city maps for visual analysis

purpose

The values ​​corresponding to the latitude and longitude data are clustered first, and then the clustering results are combined with the latitude and longitude data, and the different clustering results are displayed in the thermal panel on the map. type is the type result of clustering

Data Display

The clustering result judgment graph is clustered into 4 categories.

Clustering package loading

install.packages("factoextra")

Data read

data=read.csv("C:\\Users\\Administrator\\Desktop\\Rap\\annual_conc_by_monitor_20181.csv",header=T)

## 得出有效数据：筛选列数：根据确定要研究的变量指标名称只需要选择5列："Latitude","Longitude","Parameter_Name","Arithmetic_Mean","County_Name"
head(data)

subdata1<-data[,c("Latitude","Longitude","Parameter_Name","Arithmetic_Mean","State_Name")]

## 得出分析数据：筛选符合条件的数据：筛选城市、筛选研究指标
# 1.在指标State_Name中筛选城市New York, Vermont, New Hampshire, Massachusetts, Connecticut 这五个地方
# 2.在指标Parameter_Name中筛选 Ozone, Nitrogen dioxide, PM2. 5 local这三个变量值。

str(subdata1)

subdata2 <-  subset(subdata1,State_Name=c("New York","Vermont","New Hampshire","Massachusetts","Connecticut"))
subdata2$State=tolower(subdata2$State_Name)
newdata1 = subdata2[subdata2$Parameter_Name == 'Ozone',]
newdata2 = subdata2[subdata2$Parameter_Name == 'PM2.5 - Local Conditions',]
newdata3 = subdata2[subdata2$Parameter_Name == 'Nitrogen dioxide (NO2)',]


endata1 <- newdata1
endata2 <- newdata2
endata3 <- newdata3


#write.csv(endata1,"Ozone.csv")
#write.csv(endata2,"Nitrogen dioxide.csv")
#write.csv(endata3,"PM2.5.csv")
str(endata1)

Cluster analysis

#install.packages("fpc")
library(fpc)

## 聚类数据提取
mydata1<-endata1[,"Arithmetic_Mean"]
mydata2<-endata2[,"Arithmetic_Mean"]
mydata3<-endata3[,"Arithmetic_Mean"]

## 聚类个数判断
library(factoextra)
df1 <- scale(mydata1)
fviz_nbclust(df1, kmeans, method = "wss") + geom_vline(xintercept = 4, linetype = 2) ##得出第一个数据聚成4类

df2 <- scale(mydata2)
fviz_nbclust(df2, kmeans, method = "wss") + geom_vline(xintercept = 4, linetype = 2)

df3 <- scale(mydata3)
fviz_nbclust(df3, kmeans, method = "wss") + geom_vline(xintercept = 4, linetype = 2)

## 进行聚类

kmd1=kmeans(mydata1,centers = 4)
kmd2=kmeans(mydata2,centers = 4)
kmd3=kmeans(mydata3,centers = 4)

##聚类结果合并

cldata1<-cbind(endata1,type=kmd1$cluster) #聚类结果和原始数据合并
cldata2<-cbind(endata2,type=kmd2$cluster) #聚类结果和原始数据合并
cldata3<-cbind(endata3,type=kmd3$cluster) #聚类结果和原始数据合并

head(cldata1)
str(cldata1)
cldata1$type=factor(cldata1$type)
cldata2$type=factor(cldata2$type)
cldata3$type=factor(cldata3$type)

Map drawing

euro <- map_data("state", region = c("new york","vermont",
 "new hampshire","massachusetts", "connecticut"))

hb_map1<-merge(euro,cldata1,by.x="region",by.y="State")
hb_map2<-merge(euro,cldata2,by.x="region",by.y="State")
hb_map3<-merge(euro,cldata3,by.x="region",by.y="State")

### 地图1
ggplot(hb_map1,aes(x=long,y=lat,fill=type))+
geom_polygon(colour="grey")+
coord_map("polyconic")+ggtitle("Ozone")

### 地图2
ggplot(hb_map2,aes(x=long,y=lat,fill=type))+
geom_polygon(colour="grey")+
coord_map("polyconic")+ggtitle("PM2.5")

### 地图3
ggplot(hb_map3,aes(x=long,y=lat,fill=type))+
geom_polygon(colour="grey")+
coord_map("polyconic")+ggtitle("N02")