Search code examples
rcorrespondence-analysis

How to fix: "Error in CA(dt, graph = FALSE) : The following variables are not quantitative Var1"

I'm trying to use correspondence analysis in R. It seems like the first argument of function "CA" in FactoMineR must be a contingency table. "dt" is a contingency table, but it returns that the variables are not quantitative.

One of the levels of X1 is empty, I dont know if this is a problem in Correspondence Analysis 

library("FactoMineR")
tab1 <- table(as.factor(df$X1),as.factor(df$X2))
dt <- as.table(as.matrix(tab1))
res.ca <- CA(dt, graph = FALSE)

The output is:

Error in CA(tab1, graph = FALSE) : 
The following variables are not quantitative:  Var1
The following variables are not quantitative:  Var2
Solution

  • You have to convert your tab1 to data frame using as.data.frame.matrix(), before passing it to CA function. 

    library("FactoMineR")
tab1 <- as.data.frame.matrix(table(as.factor(df$X1),as.factor(df$X2)))
res.ca <- CA(tab1, graph = FALSE)

    Sample Data:

    df <- data.frame(X1=as.factor(c(1:101)),X2=as.factor(c(seq(0,100,1))))