I see ... here is the script that I used
=> matrice_cor_<-cor(data[,1:10], use = "everything")
=> corrplot(matrice_cor_budget, method = "number", type = "upper")
=> res<-PCA(budget, scale.unit=TRUE, ncp=5, graph = FALSE,quali.sup = c(10:10))
10 being the sample's name column
=> fviz_eig(res, choice = "variance", addlabels = TRUE)
=> corrplot(res[["var"]][["cos2"]],is.corr = FALSE)
[score plot] => plot.PCA(res, axes=c(1, 2), choix="ind", habillage = 10,
label=c("ind", "ind.sup", "quali"),new.plot=TRUE, title="score plot")
here is the probleme , it doesn't take into account the column 10 (samples)
the error displayed ( Error in plot.PCA(res, axes = c(1, 2), choix = "ind", habillage = 1, label = c("ind", :
The variable Echantillon is not quantitative)
My dataset include 9 quantitative variables ( elements : Ca, Fe, Si, Li, Ba ... + Distance from the mineralization) the 10th column represents the sample name for each row ( 400 row by sample).