19.07.2014 Views

Contents - Student subdomain for University of Bath

Contents - Student subdomain for University of Bath

Contents - Student subdomain for University of Bath

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

7.2. INTEGRATION OF RATIONAL EXPRESSIONS 171<br />

7.2 Integration <strong>of</strong> Rational Expressions<br />

The integration <strong>of</strong> polynomials is trivial:<br />

∫<br />

n<br />

∑<br />

i=0<br />

a i x i =<br />

Since any rational expression f(x) ∈ K(x) can be written as<br />

n∑<br />

i=0<br />

1<br />

i + 1 a ix i+1 . (7.4)<br />

f = p + q r with { p, q, r ∈ K[x]<br />

deg(q) < deg(r)<br />

, (7.5)<br />

and p is always integrable by (7.4), we have proved the following (trivial) result:<br />

we will see later that its generalisations, Lemmas 10 and 11, are not quite so<br />

trivial.<br />

Proposition 55 (Decomposition Lemma (rational expressions)) In the<br />

notation <strong>of</strong> (7.5), f is integrable if, and only if, q/r is.<br />

q/r with deg(q) < deg(r) is generally termed a proper rational function, but,<br />

since we are concerned with the algebaric <strong>for</strong>m <strong>of</strong> expressions in this chapter,<br />

we will say “proper rational expression.”.<br />

7.2.1 Integration <strong>of</strong> Proper Rational Expressions<br />

In fact, the integration <strong>of</strong> proper rational expressions is conceptually trivial (we<br />

may as well assume r is monic, absorbing any constant factor in q):<br />

1. per<strong>for</strong>m a square-free decomposition (Definition 32) <strong>of</strong> r = ∏ n<br />

i=1 ri i ;<br />

2. factorize each r i completely, as r i (x) = ∏ n i<br />

j=1 (x − α i,j);<br />

3. per<strong>for</strong>m a partial fraction decomposition <strong>of</strong> q/r as<br />

q<br />

r = q<br />

∏ n<br />

i=1 ri i<br />

=<br />

n∑<br />

q i<br />

r i i=1 i<br />

=<br />

n∑ ∑n i<br />

i∑<br />

i=1 j=i k=1<br />

β i,j,k<br />

(x − α i,j ) k ; (7.6)<br />

4. integrate this term-by-term, obtaining<br />

∫ q<br />

r =<br />

∑<br />

n ∑n i<br />

i∑<br />

i=1 j=i k=2<br />

−β n i,j,k<br />

(k − 1)(x − α i,j ) k−1 + ∑<br />

∑n i<br />

i=1 j=i<br />

From a practical point <strong>of</strong> view, this approach has several snags:<br />

β i,j,1 log(x − α i,j ).<br />

(7.7)<br />

1. we have to factor r, and even the best algorithms from the previous chapter<br />

can be expensive;

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!