R PCA

导入数据

data(decathlon2)
data <- decathlon2[1:23, 1:10]

#PCA
res.pca <- PCA(data, graph = FALSE)

数据分析

#特征值 可解释变异比例 累计可解释变异比例
eig.val <- get_eigenvalue(res.pca)
fviz_eig(res.pca, addlabels = TRUE, ylim = c(0, 50))

#变量和主成分的联系
var <- get_pca_var(res.pca)
head(var$coord)
fviz_pca_var(res.pca, col.var = "black")
fviz_pca_var(res.pca, col.var = "contrib",
             gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"))

#变量在主成分里的代表性
corrplot(var$cos2, is.corr=FALSE)

观测值分析

#高cos2代表变量对该主成分影响大
ind <- get_pca_ind(res.pca)
fviz_pca_ind(res.pca, pointsize = "cos2", 
             pointshape = 21, fill = "#E7B800",
             repel = TRUE)