ggfacto draws readable, interactive graphs of the
principal component, correspondence and multiple correspondence analyses
of FactoMineR. Hover over a
point, and the graph shows the crosstables behind it, each deviation
from the mean coloured: the geometry is always read against the data it
summarises. The graphs are ggplot objects, to be extended
with +.
The three analyses share one workflow: the analysis,
interpret() for its axes, ggfacto() for its
graph, and hierarchical_clust() for its clusters.
install.packages("ggfacto")
library(ggfacto)
data(tea, package = "FactoMineR")
mca <- multiple_correspondence_analysis(tea, 1:18)
interpret(mca)
ggfacto(mca, tea, sup_vars = c(sex, SPC), interactive = TRUE)See the website, with the interactive graphs: https://bricenocenti.github.io/ggfacto/, and its guide, in English or en français.