SparsEX: OOP para el Manejo de Sistemas Lineales en Memoria Extendida
Abstract
This paper commentsthe storage and retrival of elements of sparse linear systems, in extended memory of computers based on the 80386 microproeessor, under DOS.
Theset of rutines, implemented TurboPascal 6.0, is based on two objects: one of them accesses the extended memory the other one menagessparse linear systems, through hash addressing.
Other algorithms have been implemented, such as the solution of linear systems using the Crout method, and the input and output of data.
The precision of the numerical values involved is user definable.
With real data types, (11 decimal digits), 1Mbyte of extended memory can bold about 76000 numbers, in a bidimensional matrix with indexes ranging from 1 to 65535.
Theset of rutines, implemented TurboPascal 6.0, is based on two objects: one of them accesses the extended memory the other one menagessparse linear systems, through hash addressing.
Other algorithms have been implemented, such as the solution of linear systems using the Crout method, and the input and output of data.
The precision of the numerical values involved is user definable.
With real data types, (11 decimal digits), 1Mbyte of extended memory can bold about 76000 numbers, in a bidimensional matrix with indexes ranging from 1 to 65535.
Full Text:
PDFAsociación Argentina de Mecánica Computacional
Güemes 3450
S3000GLN Santa Fe, Argentina
Phone: 54-342-4511594 / 4511595 Int. 1006
Fax: 54-342-4511169
E-mail: amca(at)santafe-conicet.gov.ar
ISSN 2591-3522