Correlation curve graph and correlation heat map

# 加载 readxl 包
library(readxl)

# 读取 xlsx 文件
df<-read.csv("Seed_Data.csv")
df <- read_excel("XG.xlsx")
df <- read_excel("random_forest.xlsx",sheet = 'line_XG2')
#（更正一下，先加载包再读入表格数据，其中read后面是下划线，不是点点）
head(df)

#install.packages("GGally")
library("GGally")
#ggpairs(df,columns = 1:21) 
#ggpairs(df,columns = 1:7) 
ggpairs(df,columns = 1:8) 
df$target<-factor(df$target)
cols<-c("steelblue","yellowgreen","violetred1")
p<-ggpairs(df,
           columns = 1:8,
           aes(color=target))+
  scale_color_manual(values = cols)+
  scale_fill_manual(values = cols)+
  theme_bw()+
  theme(axis.text = element_text(colour = "black",
                                 size = 11),
        strip.background = element_rect(colour = "white",
                                        size=12),
        strip.text = element_text(face="bold"))
print(p)http://127.0.0.1:12819/graphics/plot_zoom_png?width=1463&height=844



#........................相关性分析..........
# 加载 readxl 包
library(readxl)

# 读取 xlsx 文件
mydata <- read_excel("random_forest.xlsx",sheet = 'blue_XG2')
library('corrplot')
#mydata <- read.xlsx("MMMM.xlsx",sheet="Sheet2")#读取表格MMMM中的sheet2里面的数据

View(mydata)# 这个就是看看你导入的数据，欣赏一下，不想看的话直接删除这一行就行
mydata <-as.matrix(mydata)
corr<-cor(mydata)
corrplot(corr)
corrplot(corr,tl.col = 'black')
corrplot(corr,tl.col = 'black',order = 'hclust')

corrl<-cor.mtest(mydata)#组内相关性分析，有了这一步可以在后续设置显著性水平corrl$p# 把P值调出来，热图上面标的星号按照P值这个原则
corrplot(corr,tl.col = 'black',order = 'hclust',
         p.mat = corrl$p, insig = 'blank')#对图像进行聚类处理，把显著相关的显示，其余的不要了

corrplot(corr,tl.col = 'black',order = 'hclust',
         p.mat = corrl$p,insig = 'label_sig',sig.level = c(0.001,0.01,0.05),
         pch.cex = 1,pch.col = 'red',type = 'upper')
# ,type = 'upper'把这个删除就是正方形了，沿着主轴对称，这里解释一下，正方形和三角形说明的问题都一样，正方形就是对称的

corrplot.mixed(corr,tl.col = 'black',order = 'hclust',tl.pos = "lt",diag = 'l',
               p.mat = corrl$p,insig = 'label_sig',sig.level = c(0.001,0.01,0.05),
               pch.cex = 1,pch.col = 'red')
#这个代码的意思是左下角显示相关性值的大小，右上显示图形的大小。这幅图从数值大小。
# 颜色深浅，圆圈大小都用来说明相关性这一个问题，其实都在说明一个问题
#换句话说，这三个解释相关性的方式是相互独立的，你单独拿出来哪一个都能说明问题，图形的丰富表达形式，只不过满足不同读者的阅读习惯和需求。

# 最后总结，相关性热图比较好做，复制上述代码，导入文件后就能按照需求，选择一个合适的图
corrplot.mixed(corr,tl.col = 'black',order = 'hclust',tl.pos = "lt",diag = 'l',
               p.mat = corrl$p,insig = 'label_sig',sig.level = c(0.001,0.01,0.05),
               pch.cex = 1,pch.col = 'black',insig = 'blank')

corrplot(corr,tl.col = 'black',order = 'hclust',
         p.mat = corrl$p, insig = 'blank')#对图像进行聚类处理，把显著相关的显示，其余的不要了

png(filename = "昆虫蓝色聚类2.png",width = 3000,
    
    height = 3000,units = "px",bg="white",res=300)#创作画布

corrplot(corr,tl.col = 'black',order = 'hclust',
         p.mat = corrl$p, insig = 'blank')#拓印画布

dev.off()




corrplot(corr,tl.col = 'black',order = 'hclust',
         p.mat = corrl$p, insig = 'blank',type = 'upper')#对图像进行聚类处理，把显著相关的显示，其余的不要了

png(filename = "昆虫蓝色聚类3.png",width = 3000,
    
    height = 3000,units = "px",bg="white",res=300)#创作画布

corrplot(corr,tl.col = 'black',order = 'hclust',
         p.mat = corrl$p, insig = 'blank',type = 'upper')#拓印画布

dev.off()

Data sorting format:

Figure 1-1 Correlation curve data organization

 Figure 1-2 Correlation curve data organization

Figure 1-3 into a picture 

Figure 2-1 Correlation heat map data organization format 

 Figure 2-2 Correlation heat map data organization form 

 Figure 2-3 into the picture