[R]主成分分析を行うスクリプト

R言語で主成分分析を行うスクリプト

07_PCA.R

#-------Making data------------------------------ 
data_iris<-iris[,1:4]

#-------Get and save PCA result-----------------
#PCAを行う
#center = TRUE 平均を0に正規化する
#scale = TRUE 分散は１に正規化する
pca_result<-prcomp(data_iris, center = TRUE, scale = TRUE)

#主成分得点
pca_score <- pca_result$x
write.csv(pca_score, "07_output_pca_score.csv", quote=FALSE, row.names=FALSE)

#固有ベクトル
eigenvector <- as.data.frame(pca_result$rotation)
rownames(eigenvector)<-colnames(data_iris)
write.csv(eigenvector, "07_output_eigenvector.csv", quote=FALSE)

#標準偏差、寄与率(分散比)、累積寄与率(分散比の累積値)
pca_summary<-NULL

standard_deviation<-pca_result$sdev
pca_summary<-cbind(pca_summary,standard_deviation)
  
proportion_of_Variance<-standard_deviation^2/sum(standard_deviation^2)
pca_summary<-cbind(pca_summary,proportion_of_Variance)

cumulative_Proportion <-cumsum(proportion_of_Variance)
pca_summary<-as.data.frame(cbind(pca_summary,cumulative_Proportion))

rownames(pca_summary)<-colnames(pca_score )
write.csv(pca_summary, "07_output_summary.csv", quote=FALSE)

上記の主成分分析の結果をプロットするプログラムも載せておきます。

07_PCA.R

#--------------------------------------Plot----------------------------------------
ScatterplotMatrix <- function(DataMatrix,color) {
  panel.cor<-function(x,y,digits=2,prefix="",cex.cor,...){
    usr<-par("usr")
    on.exit(par(usr))
    par(usr=c(0,1,0,1))
    r<-cor(x, y)
    txt<-format(c(r,0.123456789),digits=2)[1]
    if(missing(cex.cor))
      cex.cor<-0.8/strwidth(txt)
    text(0.5,0.5,txt,cex=1.5)
  }
  
  #Set histogram panel
  panel.hist.density<-function(x,...) {
    usr <- par("usr")
    on.exit(par(usr))
    par(usr = c(usr[1:2], 0, 1.5) )
    h <- hist(x, plot = FALSE)
    breaks <- h$breaks
    nB <- length(breaks)
    y <- h$counts
    y <- y/max(y)
    
    #draw density
    tryd <- try( d <- density(x,na.rm=TRUE,bw="nrd",adjust=1.2),silent=TRUE)
    if(class(tryd) != "try-error") {
      d$y <- d$y/max(d$y)
      lines(d,col="blue",lwd=1.5)
    }
  }
  
  
  #Set lower panel
  panel.lower <- function(x, y, ...){
    points(x, y, pch = 21, col = color)
  }
  
  pairs(DataMatrix,upper.panel=panel.cor,diag.panel=panel.hist.density,lower.panel=panel.lower)
  
  return
}

png("07_output1.png") #Open device
ScatterplotMatrix(pca_score,iris$Species)
dev.off() #Close device

png("07_output2.png") #Open device
plot(pca_summary$cumulative_Proportion)
dev.off() #Close device

結果