--- title: Multiple Correspondence Analysis (ACM) with FactoMineR (hobbies data) author: "François Husson" output: pdf_document: fig_height: 6 keep_tex: yes html_document: default word_document: default --- Script and outputs for the video of the course on the hobbies example. # Load FactoMineR ```{r} library(FactoMineR) ``` # Reading the data from a file ```{r echo=FALSE} data(hobbies) ``` ```{r eval=FALSE} hobbies <- read.table("Data_MCA_hobbies.csv", header=TRUE,sep=";") ``` ```{r} summary(hobbies) ``` # Converting the TV variable as a factor ```{r} hobbies[,"TV"] = as.factor(hobbies[,"TV"]) ``` # MCA ```{r} res.mca <- MCA(hobbies,quali.sup=19:22,quanti.sup=23,graph=FALSE) ``` # Summarizing the results on the first 2 components ```{r} summary(res.mca, ncp=2) ``` # Describing the dimensions ```{r} dimdesc(res.mca, axes=1:2) ``` # Graphic representations ```{r} plot(res.mca,invisible=c("var","quali.sup"),cex=.5,label="none") plot(res.mca,invisible=c("ind","var"),hab="quali", palette=palette(c("blue","orange","darkgreen","black","red"))) plot(res.mca,choix="var") plot(res.mca,choix="quanti.sup") barplot(res.mca$eig[,2],main="Eigenvalues", names.arg=1:nrow(res.mca$eig)) ```