|
This manual is intended to provide guidance towards the preparation
and testing of elementary function subroutines for non-vector oriented
digital computers. We believe it will be useful to systems programmers,
teachers, students of numerical analysis, hobbyists and anyone else
concerned or curious about how the elementary functions might be
computed. The functions covered are the usual assortment of algebraic,
trigonometric, and transcendental functions of real argument, including
those required by algebraic languages such as Fortran and Basic. Where
programs are included, they are written in Fortran.
Fortran also influences the terminology and notation used: We
frequently use the symbol * to mean multiplication, use ** to represent
exponentiation, and talk about the SORT program, for example. But the
ideas, algorithms, and programs presented are more widely applicable.
Many of the algorithms and accompanying test programs have been
implemented in Basic and PL/I by the authors and their students, and
some have even been implemented on programmable hand calculators.
Nevertheless, the authors make no warranty of any kind with regard to
the material in this manual. It is the reader's responsibility to
verify that the material is both correct and appropriate for his
intended usage. |