Lab 13. PCA and MCA

Assignment link

1

gospels <- read.csv("https://goo.gl/mdBVVe")
row.names(gospels) <- gospels$word
PCA <- prcomp(gospels[,2:5])

summary(PCA)

## Importance of components%s:
##                             PC1      PC2       PC3       PC4
## Standard deviation     0.003482 0.001599 0.0008725 0.0007001
## Proportion of Variance 0.761020 0.160420 0.0477900 0.0307700
## Cumulative Proportion  0.761020 0.921440 0.9692300 1.0000000

autoplot(PCA,
         shape = FALSE,
         loadings = TRUE,
         label = TRUE,
         loadings.label = TRUE)+
  theme_bw()

predict(PCA, data.frame(John = 0.05, Luke = 0.01, Mark = 0.02, Matthew = 0.02))

##              PC1         PC2         PC3         PC4
## [1,] -0.05281351 -0.01686337 0.006094607 0.002708408

2.

lev <- read.csv("http://hsequantling.wikispaces.com/file/view/Levshina_DutchCausCx.txt/567799791/Levshina_DutchCausCx.txt", sep = "\t")
MCA <- mca(lev[,-1])
lev <- cbind(lev, MCA$rs)
lev %>% 
  ggplot(aes(`1`, `2`, color = Aux))+
  geom_point()+
  stat_ellipse()+
  theme_bw()