AN INTRODUCTION TO MATRIX PASCAL: A PASCAL EXTENSION FOR SCIENTIFIC COMPUTATION

Abstract

To be a useful tool in scientific computation a programming language should provide an operator concept for the fundamental numeric data types of scientific computation. Such types include real and complex numbers, real and complex intervals as well as vectors and matrices over all of these. To this end PASCAL has been extended by means of a universal operator concept as well as a concept for functions with arbitrary result type. New standard types, operators, functions and procedures have been added to the language. In this paper we introduce only those parts of the extended language which are needed to employ and understand the operator and function concept for structured data types. In part A we introduce the new standard data types. In part B we deal with the evaluation of expressions and present the new standard operators. An introduction to the operator declaration is given in part C. We then describe the universal operator concept which allows the user to define his own operators and functions appropriate to his particular problem. The concept of maximum accuracy is extended to the evaluation of expressions containing general scalar products. The corresponding operands are presented in part E. We finally consider the problem to achieve maximum accuracy in evaluating the standard functions for all scalar types.