Graphics Reference
In-Depth Information
Both of the following inner nodes use a linoleic variable to further separate Sar-
dinian oils fromnorthern parts of Italy and on the right-hand side Apulian oils from
Sicily and Calabria. Further splits are no longer visible in this projection because
they feature other variables and are thus parallel to the visible plane. Nevertheless, it
is possible to analyze such subsequent splits, especially in an interactive context, by
a succession of sectioned scatterplots using drill-down techniques, following a few
basic rules.
Sectioned scatterplots should be preferably created using variables that are adja-
cent in the tree, that is, using split variables of nodes that are connected by an edge.
his ensures that both splits are visible in the projection.
Also, the plotted data should be restricted to the data falling in the node closer
to the root. In Fig.
.
we have used the entire dataset, since we were interested
in showing splits of the root node and its children. Figure
.
presents a sectioned
scatterplot based on data further down the tree, namely, the partition in the bottom-
right part of the scatterplot in Fig.
.
.
Sectioned scatterplots are useful for investigating the vicinity of a cut point. Some
cut points are placed in a noisy area, while others are much more clear, as is illus-
trated in Fig.
.
. However, they cannot capture more than two covariates used in
subsequent splits and thus remain suitable mainly for local analysis of a tree model.
In an interactive setting, however, it is possible to quickly “drill-down” fromFig.
.
to Fig.
.
. Linked highlighting further helps to retain the context, especially with
the help of a linked hierarchical view.
Figure
.
.
Drill-down using a sectioned scatterplot based on a subgroup induced by the tree model
