Authors:
Jaqueline A. Silveira
;
Sallles V. G. Magalhães
;
Marcus V. A. Andrade
and
Vinicius S. Conceição
Affiliation:
Universidade Federal de Viçosa (UFV), Brazil
Keyword(s):
External Memory Processing, GIS, External Algorithms.
Related
Ontology
Subjects/Areas/Topics:
Databases and Information Systems Integration
;
Enterprise Information Systems
;
Geographical Information Systems
;
Human-Computer Interaction
;
Performance Evaluation and Benchmarking
Abstract:
This paper presents a new library, named TiledMatrix, to support the development of applications that process
large matrices stored in external memory. The library is based on some strategies similar to cache memory management
and its basic purpose is to allow that an application, originally designed for
internal memory processing, can be easily adapted for external memory. It provides an interface for external memory
access that is similar to the traditional method to access a matrix.
The TiledMatrix was implemented and tested in some applications that require intensive matrix processing such as:
computing the transposed matrix and the computation of viewshed and flow accumulation on terrains represented by elevation matrix.
These application were implemented in two versions: one using TiledMatrix and another one using the Segment
library that is included in GRASS, an open source GIS.
They were executed on many datasets with different sizes and, according the tests, all app
lications ran
faster using TiledMatrix than Segment.
In average, they were 7 times faster with TiledMatrix and,
in some cases, more than 18 times faster. Notice that processing large matrices (in external memory)
can take hours and, thus, this improvement is very significant.
(More)