rel="nofollow" href="#ulink_534d2138-ace4-5f23-8c76-cd6ca840ba47">Eq. (2.84).
Figure 2.4 Variation of value of n‐term arithmetic series, Sn, normalized by first term, u0, as a function of n – for selected values of increment, k, normalized also by u0.
2.1.2 Geometric Series
A geometric series is said to exist when each term is obtained from the previous one via multiplication by a constant parameter, k, viz.
– and is said to possess ratio k and first term u0; after factoring u0 out, Eq. (2.86) becomes
An alternative form of calculating
where the first term of the summation was made apparent; multiplication of both sides by k unfolds
(2.89)
where straightforward algebraic rearrangement leads to
Ordered subtraction of Eq. (2.90) from Eq. (2.88) generates
(2.91)
where
upon canceling out symmetrical terms in the right‐hand side, and dividing both sides by 1 − k afterward, Eq. (2.92) becomes
Note that k = 1 turns nil both numerator and denominator of Eq. (2.93), so it is a (common) root thereof; Ruffini’s rule (see below) then permits reformulation of Eq. (2.93) to
(2.94)
that mimics Eq. (2.87), as expected. Revisiting Eq. (2.72) with division of both sides by u0, one may insert Eq. (2.93) to get
Eq. (2.95) is illustrated in Fig. 2.5. Upon inspection of this plot, one anticipates a horizontal asymptote for k = 0.5 (besides k = 0); in general, one indeed finds that
(2.96)
at the expense of Eqs. (2.72), (2.73), and (2.95), which is equivalent to
– and reduces to merely
should k lie between −1 and 1, as it would give rise to a convergent series (i.e. kn + 1 → 0 when n → ∞, in this case). All remaining values, i.e. k ≤ −1 or k ≥ 1, produce divergent series – as apparent in Fig. 2.5; when k is nil, Eq. (2.93) turns to just
(2.99)
which corresponds to the trivial case of only the first term of said series being significant – and fully consistent with Eqs. (2.97) and (2.98).
Figure 2.5 Variation of value of n‐term geometric series, Sn, normalized by first term, u0, as a function of n – for selected values of ratio, k.
An alternative proof of Eq. (2.93) comes from finite induction; one should first confirm that it applies to n = 0, i.e.
(2.100)