Correspondence analysis simple example

Simple CA

# Ca fun
# from the qualtrics example https://www.qualtrics.com/eng/correspondence-analysis-what-is-it-and-how-can-i-use-it-to-measure-my-brand-part-1-of-2/
library(ca)

# this just re-creates their initial table

mydata <- data.frame(
  Brand = c(rep('Butterbeer',3),
            rep('Squishee', 3),
            rep('Slurm', 3),
            rep('Fizzy Lifting Drink', 3),
            rep('Brawndo', 3)),
  Attributes = rep(c('Tasty', 'Aesthetic', 'Economic'), 5),

Counts = c(5,7,2,18,46,20,19,29,39,12,40,49,3,7,16)
)
mytable <- xtabs(Counts ~ ., mydata)

# Brand on rows

# Actual 'work' starts heere

prop.table(mytable, 1) # row percentages
prop.table(mytable, 2) # column percentages
fit <- ca(mytable)
print(fit) # basic results
summary(fit) # extended results
#Principal Normalization Correspondence Analysis (Symmetrical, French Map, Canonical)
plot(fit) # symmetric map
plot(fit, mass = TRUE, contrib = "absolute", map =
       "rowgreen", arrows = c(FALSE, TRUE)) # asymmetric map