Graphics Reference
In-Depth Information
run again (if necessary), and new results updating the current views automatically
appear.
Let us start the analysis results interpretation by looking at the representation of
variables with respect to the first two factors. In Fig.
.
we see the first factorial
plan whose total associated inertia is
.
%. We observe all indicators, but YT con-
tributes to the first factor. With respect to the second factor, instead, they split into
two groups.
Aterthefactorialaxesareinterpretedaccordingtothevariablesintheanalysis,we
shit our attention to the statistical units' configuration on the same factorial space.
InFig.
.
weobserve theconfiguration ofpoints onthe firstfactorial plan.Asinthe
mostofdata-miningapplications,wenoticethatthelargestpartofthestatisticalunits
is massed on the axis origins. his happens because the analysis is strongly affected
by a few influence points.
Inordertodetectwhichkindofunitshavethegreatestinfluence,thesystemallows
the user to differentiate points according to the modalities of a categorical variable.
In this example, on the right-hand side of Fig.
.
points are colored according to
the modalities of variable Q (working status): blue points correspond to “employed”,
cyan indicates “self-employed”, and black indicates “unemployed”.
Figure
.
.
Representation of variables: SHIW data
