A configurable environment for the visualization of data
Benoı̂t Ozell · 1996
The characteristic elements of a data visualization environment are its functionality and its dialogue. Because of the diversity in numerical simulation types as well as in various interest in visualization and analysis, the tendency has been to use data visualization programs adapted to their needs and field of research. Despite the name and signification attached to each solution entity, the basic functions of data visualization are, however, the same; i.e. information extraction and representation. The objective is to present a methodology for the building of a configurable software program for visualization and analysis. The design of a field-specific or even project-specific environment is possible, while maintaining a core program to perform all the work of data extraction, manipulation and rendering. To achieve this goal, it is necessary to interpret visualization data in a generic sense. The grid, the primary variables of the problem, and the derived variables are all considered as a number of simple scalar fields. Basic entities in the program then provide building blocks to create images. By combining these in specific ways, one can adapt or configure the appearance of the interface to reflect the requirements of the user. This allows for a complete variety of associating the data to the various representations possible. A visualization and analysis language is presented on the basis of a semiological analysis of the visualization process. Its entities are defined in a paradigm while the possible combinations of these are described in its syntagm. This language allows for the specification of data, the associated representation, the probe positions and the displaying modes. The data dictionary or paradigm of this methodology includes: the grid, the solution, the support, the travel path, the entity, the tracer, the script, the window, the figure and the expression. As well as the language of algebra allows the definition of operations acting on numbers, this algebra defines a syntagm of functions on data extraction and presentation. There are four different modes resulting from the association of the program basic entities: (1) Static: Item Set + Support + Windows; (2) Dynamic: Item Set + Travel Path + Windows; (3) Injection: Tracer + Support + Windows; (4) Animation: Script + Windows. The expressions, directly interpreted by the program gives a tool for the creation of derived fields from the solution information. It is therefore possible, for example, to calculate secondary variables, deduct quality measures of a grid or a solution, as well as producing error estimators.