R繪圖（01）——ggplot畫柱狀圖和多圖拼接

數據集形式：

 

目標構想：

需要使用到的變量為ICEWSGOVT_bin（GV）, ICEWSNGOVT_bin（NGV）,Year_F（Y）,GrowingSeason_Month（GSM），共4個變量。其中第一和第二變量為1和0的名義變量，需要完成的事情是，根據GSM作為分類條件，統計隨着Y的改變，GV和NGV的為1的個數。

因為由兩個數據庫，每個數據庫包含GSM=1和GSM=0的兩種條件，故而總共存在4張圖片，隨后需要將4張圖片排列在一張圖片上。

 

實現路徑（我的菜鳥思維）：

1.從繪畫一個圖開始，繪畫柱狀圖需要准備好表格數據，所以應該准備好繪圖的數據，繪畫一張直方圖，目標數據格式如圖

 

 2.為了准備這個表格，就需要根據條件篩選數據，再匯總數據整理成為表格，這就涉及到分類、分組計算。所以目標是將大數據庫整理成目標表格的格式。

3.獲得數據庫之后，就可以對圖進行美化和調整，因為不設置各類參數的話，默認輸出圖片不太符合主流投稿或者寫文的規范（例如灰色底色，沒有坐標軸等）。

3.繪畫好了一張圖，就可以效仿繪制剩下的圖，最后排版。

 

最終實現效果如圖：

 

 

具體實現的代碼（無數據路徑版本）：

library(reshape2)
ICE<-Data_ICEWS_Trim
ACL<-Data_ACLED_Trim

#######准備目標表格的數據#####
#######data base - ICE ######
var<-c("ICEWSGOVT_bin","ICEWSNGOVT_bin","Year_F","GrowingSeason_Month")
new<-ICE[var]
colnames(new) <- c("GOVT","NGOVT","YEAR","GSM")
sums_1 <- matrix(NA, nrow=8, ncol=3)
sums_0 <- matrix(NA, nrow=8, ncol=3)
for (i in 1:8){
  sums_1[i,1] <- i+2010
  sums_1[i,2] <- sum(new$GOVT[new$YEAR==i+2010&new$GSM==1])
  sums_1[i,3] <- sum(new$NGOVT[new$YEAR==i+2010&new$GSM==1])
}
for (i in 1:8){
  sums_0[i,1] <- i+2010
  sums_0[i,2] <- sum(new$GOVT[new$YEAR==i+2010&new$GSM==0])
  sums_0[i,3] <- sum(new$NGOVT[new$YEAR==i+2010&new$GSM==0])
}
sums_1<-as.data.frame(sums_1)
sums_0<-as.data.frame(sums_0)
colnames(sums_1)<-c("YEAR","GOV","NGOV")
colnames(sums_0)<-c("YEAR","GOV","NGOV")
varc<-c("YEAR")
rs_data1<- melt(sums_1, id.vars = varc)
rs_data0<- melt(sums_0, id.vars = varc)
###### data base - ACL ######
var2<-c("acledGOVT_bin","acledNGOVT_bin","Year_F","GrowingSeason_Month")
new<-ACL[var2]
colnames(new) <- c("GOVT","NGOVT","YEAR","GSM")
Asums_1 <- matrix(NA, nrow=3, ncol=3)
Asums_0 <- matrix(NA, nrow=3, ncol=3)
for (i in 1:3){
  Asums_1[i,1] <- i+2016
  Asums_1[i,2] <- sum(new$GOVT[new$YEAR==i+2016&new$GSM==1])
  Asums_1[i,3] <- sum(new$NGOVT[new$YEAR==i+2016&new$GSM==1])
}
for (i in 1:3){
  Asums_0[i,1] <- i+2016
  Asums_0[i,2] <- sum(new$GOVT[new$YEAR==i+2016&new$GSM==0])
  Asums_0[i,3] <- sum(new$NGOVT[new$YEAR==i+2016&new$GSM==0])
}
Asums_1<-as.data.frame(Asums_1)
Asums_0<-as.data.frame(Asums_0)
colnames(Asums_1)<-c("YEAR","GOV","NGOV")
colnames(Asums_0)<-c("YEAR","GOV","NGOV")
Ars_data1<- melt(Asums_1, id.vars = varc)
Ars_data0<- melt(Asums_0, id.vars = varc)



#####繪圖#####
library(ggplot2)
library(ggpubr)
ICE_1<-ggplot(rs_data1,aes(x=YEAR,y=value,fill=variable))+geom_bar(position="dodge",stat="identity")
ICE_1<-ICE_1+ylab("Number")+labs(fill="Type",subtitle = "GSM=1")
ICE_1<-ICE_1+ theme_bw()+theme(panel.border = element_blank(),panel.grid.major = element_blank(),panel.grid.minor =element_blank(),axis.line = element_line(colour = "Black"))
ICE_1<-ICE_1+scale_y_continuous(limits =c(0, 200) ,expand = c(0,0))+scale_x_continuous(breaks = seq(2011, 2018, by = 1))

ICE_0<-ggplot(rs_data0,aes(x=YEAR,y=value,fill=variable))+geom_bar(position="dodge",stat="identity")
ICE_0<-ICE_0+ylab("Number")+labs(fill="Type",subtitle = "GSM=0")
ICE_0<-ICE_0+ theme_bw()+theme(panel.border = element_blank(),panel.grid.major = element_blank(),panel.grid.minor =element_blank(),axis.line = element_line(colour = "Black"))
ICE_0<-ICE_0+scale_y_continuous(limits =c(0, 200) ,expand = c(0,0))+scale_x_continuous(breaks = seq(2011, 2018, by = 1))

ACL_1<-ggplot(Ars_data1,aes(x=YEAR,y=value,fill=variable))+geom_bar(position="dodge",stat="identity")
ACL_1<-ACL_1+ylab("Number")+labs(fill="Type",subtitle = "GSM=1")
ACL_1<-ACL_1+ theme_bw()+theme(panel.border = element_blank(),panel.grid.major = element_blank(),panel.grid.minor =element_blank(),axis.line = element_line(colour = "Black"))
ACL_1<-ACL_1+scale_y_continuous(limits =c(0, 700) ,expand = c(0,0))+scale_x_continuous(breaks = seq(2017, 2019, by = 1))

ACL_0<-ggplot(Ars_data0,aes(x=YEAR,y=value,fill=variable))+geom_bar(position="dodge",stat="identity")
ACL_0<-ACL_0+ylab("Number")+labs(fill="Type",subtitle = "GSM=0")
ACL_0<-ACL_0+ theme_bw()+theme(panel.border = element_blank(),panel.grid.major = element_blank(),panel.grid.minor =element_blank(),axis.line = element_line(colour = "Black"))
ACL_0<-ACL_0+scale_y_continuous(limits =c(0, 700) ,expand = c(0,0))+scale_x_continuous(breaks = seq(2017, 2019, by = 1))

ggarrange(ICE_1, ICE_0, ACL_1, ACL_0, labels = c("A", "B", "C","D"),ncol = 2, nrow = 2)