\documentclass{article}
\usepackage{axiom}
\begin{document}
\title{\$SPAD/src/input pfr.input}
\author{The Axiom Team}
\maketitle
\begin{abstract}
\end{abstract}
\eject
\tableofcontents
\eject
\section{Partial Fractions}
A {\it partial fraction} is a decomposition of a quotient into a
sum of quotients where the denominators of the summands are
powers of primes.
For example, the rational number $1 \over 6$ can be decomposed
into ${{1 \over 2} + {-1 \over 3}}.$ It is possible to compute
partial fractions of quotients of objects belonging to domains
belonging to the category {\sf EuclideanDomain}.
For example, {\sf Integer, Gaussian(Integer)} and {\sf
UnivariatePoly(x, RationalNumber)} all belong to {\sf
EuclideanDomain}.
In the examples below, we'll demonstrate how to decompose
quotients of each of these kind of objects into partial fractions.

It is necessary that we know how to factor the denominator when
we want to compute a partial fraction.  Although the interpreter
can often do this automatically, it is easier if you just include
a call to {\em factor}.  The main function for computing partial
fractions is called {\em partialFraction} and we'll use this to
compute a decomposition of $1 \over {10!}$.
The first argument to {\em partialFraction} is the numerator of
the quotient and the second argument is the factored denominator.
<<code>>=
partialFraction(1,factor factorial 10)
@
Since the denominators are powers of primes, it may be possible to
further expand the numerators with respect to those primes. Use the
function {\em padicFraction} to do this.  The function
{\em compactFraction} returns an expanded fraction into the usual
form.  The compacted version is used internally for computational
efficiency.
<<code>>=
f := padicFraction %
compactFraction %
@
Given a partial fraction, it is possible to add, subtract, multiply
and divide them.  In addition, it is possible to extract the parts
of the decomposition.  The number of terms in fractional part is
computed by {\em numberOfFractionalTerms}.
<<code>>=
numberOfFractionalTerms f
@
This does {\it not} include the whole part of the fraction, though
it can be gotten by calling {\em wholePart}.
<<code>>=
wholePart f
@
Individual terms can be extracted with {\em nthFractionalTerm}.
<<code>>=
t3 := nthFractionalTerm(f,3)
@
The numerator and denominator of the first fractional term are gotten
by calling {\em firstNumer} and {\em firstDenom}, respectively.
<<code>>=
firstNumer t3
firstDenom t3
@
Given two gaussian integers (see \S \ref{gaussz}), it is possible
to decompose their quotient into a partial fraction.
<<code>>=
g := - 13 + 14 * %i
1/g
partialFraction(1,factor g)
@
To convert back to a quotient, simply use a coercion.
<<code>>=
% :: FRAC COMPLEX INT
@
Also, since {\sf QF G I} is {\it isomorphic} to
{\sf G RN} is further possible to
coerce the last result to a gaussian with rational number real and
imaginary parts.
<<code>>=
% :: COMPLEX FRAC INT

)clear all
@
To conclude this section, we'll compute the decomposition of
$$
1 \over {\prod_{i=1}^{4} {{(x+1)}^i}}
$$
The polynomials in this object will be of type
{\sf UnivariatePoly(x, RationalNumber)}, though it is easier to write
as {\sf P[x] RN}.
We'll use the {\em prfac} function (see \S \ref{factobj}) to
directly compute the denominator in factored form.
<<code>>=
u : FR UP(x,FRAC INT) := reduce(*,[primeFactor(x+i,i) for i in 0..4])
@
The compact and expanded partial fractions for the quotient are
<<code>>=
partialFraction(1,u)
padicFraction %
@
\section{License}
<<license>>=
--Copyright The Numerical Algorithms Group Limited 1991.
@
<<*>>=
<<license>>

)set out len 57
)time off
)clear all

<<code>>
@
\eject
\begin{thebibliography}{99}
\bibitem{1} nothing
\end{thebibliography}
\end{document}