Quotients by Normalizing the Denominator

Why Normalize the Denominator?

A geometric series is a useful starting point for building power series for rational functions. The central identity is \(1/(1-u)=\sum_{n=0}^{\infty}u^n\), valid when \(|u|<1\). A denominator that does not already have the form \(1-u\) can often be rewritten to have it, at the cost of factoring out a constant.

After that normalization, the reciprocal has a geometric expansion. Multiplying by the numerator then gives a series for the quotient. The key tasks are to keep track of the factored constant, state the region where the expansion converges, and multiply the series correctly.

The Geometric Expansion After Normalization

The condition \(|h(x)|<1\) is the convergence condition for the geometric series. In particular, it is not enough to recognize a geometric-looking expression: the value of \(h(x)\) must satisfy this strict inequality at the point under consideration. Also, the factor \(1/C\) remains in the reciprocal. Omitting it changes every coefficient.

The geometric identity itself follows from the finite geometric sum. For \(u\ne 1\), the partial sum through the \(N\)th power satisfies the following identity:

If \(|u|<1\), then \(u^{N+1}\) tends to zero as \(N\) tends to infinity. Taking the limit of both sides gives \(\sum_{n=0}^{\infty}u^n=1/(1-u)\). Substituting \(u=h(x)\) proves the normalized reciprocal formula wherever \(|h(x)|<1\).

Multiplying by a Numerator

Suppose the numerator is a polynomial \(P(x)\), and the denominator is normalized as \(C(1-h(x))\). In the region where \(|h(x)|<1\), multiply the reciprocal series by \(P(x)/C\). If \(P\) is a polynomial in \(x\) and \(h(x)\) is linear in \(x\), collecting like powers of \(x\) gives a power series. The coefficients can be found by adding the finite contributions that produce each power.

For example, if \(P(x)=p_0+p_1x\) and the reciprocal series has coefficients \(a_n\), then the product has constant coefficient \(p_0a_0\); its coefficient of \(x^k\), for \(k\geq 1\), is \(p_0a_k+p_1a_{k-1}\). This is ordinary multiplication of series: a term of degree \(k\) can come only from terms whose degrees add to \(k\).

More generally, if the numerator already has a power-series representation, multiply the two series where both converge. Inside their common interval of convergence, power series converge absolutely, so their Cauchy product is valid; the coefficient of a given power is the finite sum of products whose indices add to that power. The examples below use polynomial numerators so that the coefficient calculations remain explicit.

Worked Examples

Common Pitfall: Losing the Factored Constant

A frequent error is to expand the denominator after recognizing its variable part but forget the constant that was factored out. For instance, the incorrect calculation

is missing the factor \(1/3\). The error is visible even at \(x=0\): the left side is \(2/3\), while the proposed right side is \(2\). The correct normalization is \(3-x=3(1-x/3)\), so the reciprocal is \(\frac{1}{3}\sum_{n=0}^{\infty}(x/3)^n\), and the quotient is \(\frac{2+x}{3}\sum_{n=0}^{\infty}(x/3)^n\).

A second check is to multiply the proposed series by the original denominator. If the product does not give the numerator, a factor or coefficient has been lost. It is also important not to claim that the geometric series converges wherever the rational function is defined: the expansion requires the stricter condition \(|h(x)|<1\).

Why This Method Matters

Normalizing a denominator turns many quotients into a familiar geometric expansion. It provides a direct way to obtain coefficients without first solving for them recursively, and the resulting series can be checked by multiplication. The method also makes the convergence restriction explicit: it comes from the normalized variable part of the denominator.

When applying the method, keep the steps distinct. First identify \(C\) and \(h(x)\); next solve \(|h(x)|<1\); then write the reciprocal with its factor \(1/C\); finally multiply by the numerator and collect coefficients. These checks help distinguish a correct series representation from a formally plausible but invalid one.