(Place the mouse over the members' faces)

The central idea that unifies the Group is the statistical analysis
of **compositional data**. In Geology, Petrology, Chemistry, Economics,
Archaeometry, etc. we are used to work with vectors whose components represent
relative parts of a whole. In fact, we are working with the mathematical
foundations under the statistical analysis of this kind of data, and also
with some of their wide-range applications.

From the vast workings of professor J. Aitchison in this field, built
up from a statistical point of view since the early eighties, we have seen
that rigorous mathematical foundations are needed to properly understand
and apply these techniques. We base them in the definition of a geometric
structure for the **Simplex** (the support space of compositional data).
Afterwards, traditional statistical techniques (cluster, disciminant, factor
analysis, regression models, etc.) are being adapted to this structure:
we are covering both their real-case application and theoretical development,
from measure theory or algebric geometry to differential calculus in the
simplex.

Furthermore, we are developing our own statistical package, the
CoDaPack,
due the total lack of a specific and user-friendly software intended to
treat compositional data.

Here you can view a video about the activity of the research group appeared on TV3(Televisió de Catalunya) (minute 2:08):
Compositional data of the soil of Mercury