R绘图（01）——ggplot画柱状图和多图拼接

数据集形式：

 

目标构想：

需要使用到的变量为ICEWSGOVT_bin（GV）, ICEWSNGOVT_bin（NGV）,Year_F（Y）,GrowingSeason_Month（GSM），共4个变量。其中第一和第二变量为1和0的名义变量，需要完成的事情是，根据GSM作为分类条件，统计随着Y的改变，GV和NGV的为1的个数。

因为由两个数据库，每个数据库包含GSM=1和GSM=0的两种条件，故而总共存在4张图片，随后需要将4张图片排列在一张图片上。

 

实现路径（我的菜鸟思维）：

1.从绘画一个图开始，绘画柱状图需要准备好表格数据，所以应该准备好绘图的数据，绘画一张直方图，目标数据格式如图

 

 2.为了准备这个表格，就需要根据条件筛选数据，再汇总数据整理成为表格，这就涉及到分类、分组计算。所以目标是将大数据库整理成目标表格的格式。

3.获得数据库之后，就可以对图进行美化和调整，因为不设置各类参数的话，默认输出图片不太符合主流投稿或者写文的规范（例如灰色底色，没有坐标轴等）。

3.绘画好了一张图，就可以效仿绘制剩下的图，最后排版。

 

最终实现效果如图：

 

 

具体实现的代码（无数据路径版本）：

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

#######准备目标表格的数据#####
#######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)



#####绘图#####
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)