Graphics Reference
In-Depth Information
Table
.
.
Six macroeconomic performance indicators of
OECD countries (percentage change from
the previous year, September
/
)
Country
Label
GDP
LI
UR
IR
TB
NNS
Australia
A-lia
.
.
.
.
.
.
Canada
Can
.
.
.
.
.
.
Finland
Fin
.
−
.
.
.
.
.
France
Fra
.
.
.
.
.
.
Spain
Spa
.
.
.
.
.
.
Sweden
Swe
.
.
.
.
.
.
United States
USA
.
.
.
.
−
.
.
Netherlands
Net
.
.
.
.
.
.
Greece
Gre
.
.
.
.
.
.
−
Mexico
Mex
.
.
.
.
.
.
Portugal
Por
.
−
.
.
.
−
.
.
Austria
A-tria
.
.
.
.
−
.
.
Belgium
Bel
.
.
.
.
.
.
−
Denmark
Den
.
.
.
.
.
.
Germany
Ger
.
−
.
.
.
.
.
Italy
Ita
.
−
.
.
.
.
.
Japan
Jap
.
.
.
.
.
.
Norway
Nor
.
.
.
.
.
.
Switzerland
Swi
.
.
.
.
.
.
United Kingdom
UK
.
.
.
.
−
.
.
Although we are aware that many innovative visualization methods appear every
year in thespecialized literature, wewill not discussthecapabilities ofthese methods
here;instead wewill focusourattention on those visualization methods in whichthe
concept of distance plays a central role.
Factorial Analysis
4.3
In statistics, factorial analysis (FA) refers to a set of methods that permit one to re-
ducethe dimension ofadata matrix with respecttoaleast-squares criterion (Mizuta,
). he geometric formalization of the problem was one of the keys to the great
success and quick dissemination of the methods.
Given a generic data matrix
X
of order n
p, the aim of the methods is to replace
p original variables with a set of q ordered and orthogonal factors that are obtained
as a linear combination of the original variables, where q
p. Factors are ordered
according to the information they carry. Orthogonality ensures consistent represen-
tations based on Cartesian space and allows us to split the whole variability into an
additive linear model based on independent variables.
Ofthefactorialmethods,principalcomponentanalysis(PCA)isprobablythebest,
most used, and most implemented in statistical sotware packages. PCA deals with
<
