Database Reference
In-Depth Information
known planets.
∗
A relatively recent overview in Adelmann et al. provides a
good survey of the field's breadth and depth: algorithms for visualizing scalar,
vector, and tensor fields, geometric modeling, virtual environments for visu-
alization, large data visualization, perceptual and cognitive issues in visual-
ization, visualization software and frameworks, software architecture, and so
forth.
Visualization is a very data-intensive science: visualization algorithms take
as input vast amounts of data produced by simulation or experiment, and
then transform that data into imagery. It turns out, as we shall explore in
this chapter, that visualization reveals a somewhat different view of scientific
data management challenges than are examined elsewhere in this topic. For
example, a data ordering and storage layout that works well for saving data
from memory to disk may not be the best thing for subsequent visual data
analysis algorithms.
This chapter will present four broad topic areas under this general rubric:
(1) a view of SDM-related issues from the perspective of implementing a
production-quality, parallel capable visual data analysis infrastructure; (2)
novel data storage formats for multiresolution, streaming data movement, ac-
cess and use by postprocessing tools; (3) data models, formats and APIs for
performing ecient I/O for both simulations and postprocessing tools, dis-
cussion of issues, and previous work in this space; (4) how combining state-of-
the-art techniques from scientific data management and visualization enables
visual data analysis of truly massive datasets.
9.2 Production-Level, Parallel Visualization Tool
Perspective on SDM
A production-level, parallel visualization tool is a robust program that is used
by a potentially large population of users to perform diverse visualizations
and analyses, normally on data from many different types of file formats and
with varying types of data models. As such, these tools are somewhat different
from scientific simulation codes in that
They “unify” many different data models. For example, they support
many mesh types, field types, and various centerings for those fields
(point centered, cell centered, etc).
They are not the originators of the semantics placed on the data. There-
fore, the meaning of each of the arrays of data must somehow be provided
to the visualization tool.
∗
See http://en.wikipedia.org/wiki/Orrery
