Most series cannot be summed in closed form: we can only prove that they converge. The geometric series 1+r+r2+⋯ is the great exception. Its partial sums have an exact formula, so we can see precisely when it converges and to what. That makes it the yardstick against which other series are measured (the comparison test of the last lesson needs something to compare with), and it explains a fact from school arithmetic: why recurring decimals are exactly the fractions.
The idea
Think of the partial sums as a walk along the number line. Start at 0, take a step of length 1, then a step of length r, then r2, and so on: after n steps you stand at sn=1+r+⋯+rn−1. Each step is r times the one before. If r is negative, each step also turns round.
ExploreTake the series apart
Try this.Start with r=0.5 and increase the number of terms. Then try r=0.9, r=−0.6, r=−1, r=1 and r=1.1.
1+r+r2+⋯; sn=1−r1−rn
sₖstep k
step k: from sₖ to sₖ₊₁, length rᵏ
s = 1/(1 − r)
what is left: s − sₙ
s6=1.96875. What is left is s−s6=r6/(1−r)=0.03125: the tail of the walk is the whole walk shrunk by r6. |r| < 1: the steps shrink geometrically and the walk closes in on 1/(1 − r) = 2.
What to notice
For ∣r∣<1 the purple gap s−sn is always the whole walk shrunk by the factor rn: that self-similarity is the reason the walk closes in on 1/(1−r). For negative r the walk zigzags across its limit. For r=−1 it hops between 1 and 0 for ever, and for ∣r∣>1 the steps grow and the walk runs away, to one side (r>1) or to both (r<−1).
The algebra behind the picture is one line. Multiply sn by r and subtract:
sn−rsn=(1+r+⋯+rn−1)−(r+r2+⋯+rn)=1−rn,
so, if r=1,
sn=1−r1−rn=1−r1−1−rrn.
The partial sum is a fixed number 1/(1−r) minus a multiple of rn. Everything therefore depends on what rn does, and that was settled in §72: rn→0 if ∣r∣<1, rn→1 if r=1, rn→+∞ if r>1, and rn oscillates if r≤−1 (finitely if r=−1, infinitely if r<−1).
ExploreHow fast does it converge?
Try this.This plots the error ∣s−sn∣ on a logarithmic scale. Move r between 0.1 and 0.95.
∑n=0∞rn
n|s − sₙ| (log scale)
|s − sₙ| for rn
Plot
After 30 terms s29=1.999999998; the sum is 2, so ∣s−s29∣ = 1.86 × 10⁻⁹. The series converges to 2.
What to notice
The points lie on a straight line: on a log scale, multiplying by r at every step is a constant downward step of log10(1/r). For r=0.5 you gain a decimal digit about every 3.3 terms; for r=0.9 you need about 22 terms per digit.
Precise statements
TheoremThe geometric seriesHardy: §78
The series 1+r+r2+⋯
converges, with sum 1−r1, if −1<r<1;
diverges to +∞ if r≥1;
oscillates finitely if r=−1;
oscillates infinitely if r<−1.
Combined with §77 (1), (2) and (4), this sums every series of the form arm+arm+1+⋯: for ∣r∣<1 it is arm/(1−r).
Hardy’s Examples XXIX apply the theorem to decimals. A decimal ⋅a1a2a3… (digits ak from 0 to 9) means the series a1/10+a2/102+a3/103+⋯. It always converges, because its terms are ≥0 and its partial sums are at most 9/10+9/102+⋯=1.
TheoremRecurring decimals are fractions, and converselyHardy: Examples XXIX, 2–6
A decimal that terminates or recurs (repeats a fixed block of digits for ever from some point on) represents a rational number. Conversely, every rational number between 0 and 1 has a decimal expansion that terminates or recurs. Consequently a decimal that neither terminates nor recurs represents an irrational number.
Proofs
ProofThe geometric series
Strategy
Use the exact formula sn=1−r1−rn for r=1, and read off the behaviour of sn from that of rn (§72). Treat r=1 separately.
r=1
Every term is 1, so sn=n→+∞.
−1<r<1
Here rn→0, so sn=1−r1−1−rrn→1−r1 by the limit theorems for sums and constant multiples.
r>1
Every term is at least 1, so sn≥n and sn→+∞. (Or: rn→+∞ and 1−r<0.)
r=−1
The terms are 1,−1,1,−1,…, so sn is 1 for odd n and 0 for even n: bounded, without a limit. It oscillates finitely.
r<−1
Write r=−ρ with ρ>1, so 1−r=1+ρ>0. For even n, sn=(1−ρn)/(1+ρ)→−∞; for odd n, sn=(1+ρn)/(1+ρ)→+∞. So sn is unbounded in both directions, and neither tends to +∞ nor to −∞: it oscillates infinitely.
∎
There is a tempting shortcut (Examples XXX, 7): if s=1+r+r2+⋯, then s=1+r(1+r+⋯)=1+rs, so s=1/(1−r). The algebra uses §77 (1) and (4), which apply only to a series already known to converge. Without that hypothesis the same algebra “proves” 1+2+4+⋯=−1.
ProofRecurring decimals and fractions
Strategy
One direction is a geometric series: a block of n repeating digits is a geometric series with ratio 10−n. For the other, the long division of p by q has only q possible remainders, so it must eventually repeat itself.
A pure recurring decimal is a fraction
Let the block a1a2…an repeat for ever, and let A be the integer with those digits. Then ⋅a1…an=A10−n+A10−2n+⋯, a geometric series with first term A10−n and ratio 10−n, so its sum is 1−10−nA10−n=10n−1A.
Any recurring or terminating decimal is a fraction
A decimal that starts recurring after m digits is 10−m times (an integer plus a pure recurring decimal), which is rational. A terminating decimal is a finite sum of fractions.
Long division by q has at most q remainders
Let 0<p/q<1. The digits of p/q come from long division: at each stage the remainder ρk is one of 0,1,…,q−1, and the next digit and the next remainder depend only on ρk (the digit is [10ρk/q], the new remainder is 10ρk−q[10ρk/q]).
So the digits repeat
Among the first q+1 remainders two are equal, say ρj=ρj+n with n≥1. From then on the division repeats exactly, so the digits recur with a block of length n. If some remainder is 0, the decimal terminates.
The consequence
A decimal that neither terminates nor recurs cannot be rational, since every rational has a terminating or recurring expansion, and a number has only one expansion apart from the case of trailing 9s (Examples XXIX, 6–7), which also terminates or recurs.
∎
Worked examples and exercises
Exercise●○○Hardy: Examples XXIX, 1
Express the decimal ⋅2171313…, in which the block 13 recurs for ever after the digits 217, as a fraction in lowest terms. Which theorems of §77 does your calculation use?
Hint
Split it as ⋅217 plus 13/105+13/107+13/109+⋯.
Solution
The recurring part is a geometric series with first term 13/105 and ratio 1/102:
We used: grouping the digits into the terms 13/105+2k (inserting brackets, §77 (8)); the sum of a geometric series and its multiples (§77 (4)); and putting the three terms 2/10, 1/102, 7/103 in front (§77 (1)).
Exercise●○○Hardy: Examples XXX, 3
Show that 1+2r+2r2+2r3+⋯ converges when ∣r∣<1, with sum 1−r1+r, in two ways: (α) as −1+2(1+r+r2+⋯), and (β) as 1+2(r+r2+⋯).
Hint
Each way uses §77 (1) or (2) together with (4).
Solution
(α) 1+r+r2+⋯=1/(1−r); doubling (§77 (4)) gives 2+2r+2r2+⋯=2/(1−r). Our series is this one with its first term 2 replaced by 1: deleting the first term (§77 (2)) and then putting 1 in front (§77 (1)) gives the sum 2/(1−r)−2+1=1−r2−(1−r)=1−r1+r.
(β) r+r2+⋯=r/(1−r) (§77 (2) or (4)), so 2r+2r2+⋯=2r/(1−r), and putting 1 in front gives 1+1−r2r=1−r1+r.
Exercise●●○Hardy: Examples XXX, 5
For which r does (1−r)+(r−r2)+(r2−r3)+⋯ converge, and what is its sum?
Hint
It is a telescoping series. What is sn?
Solution
The n-th partial sum is 1−rn, which tends to a limit exactly when rn does, that is, when −1<r≤1 (§72). The sum is 1 if −1<r<1, and 0 if r=1 (every term is then 0). Note the surprise: the sum is not a continuous function of r at r=1. For other r the series diverges (r>1) or oscillates (r≤−1).
Exercise●●○Hardy: Examples XXX, 8
Find when r+1+rr+(1+r)2r+⋯ converges, and its sum.
Hint
For r=−1 it is a geometric series with first term r and ratio 1/(1+r). Treat r=0 separately.
Solution
If r=−1 the terms after the first are undefined. If r=0 every term is 0 and the sum is 0. Otherwise the series is r times the geometric series with ratio q=1/(1+r), which converges if and only if ∣q∣<1, that is ∣1+r∣>1, that is r>0 or r<−2. Then the sum is
1−1+r1r=rr(1+r)=1+r.
For −2≤r<0 with r=−1, ∣q∣≥1, and r times a non-convergent series does not converge (§77 (5)).
Exercise●●○Hardy: Examples XXIX, 8
Show that ⋅101001000100001…, in which the runs of zeros between successive 1s have lengths 1,2,3,…, represents an irrational number.
Hint
By the theorem, it is enough to show that the decimal neither terminates nor recurs.
Solution
It does not terminate, since there are infinitely many 1s. Suppose it recurred with a block of length p from the m-th digit on. Far enough along there is a run of more than m+p consecutive zeros, so some complete block of p digits after position m consists only of zeros. Then every later digit is 0, contradicting the infinitely many 1s. So the decimal neither terminates nor recurs, and the number is irrational.
Check yourself
Check yourselfChoose one
What does 1−1+1−1+⋯ (the geometric series with r=−1) do?
Why
The partial sums are 1,0,1,0,…: bounded and without a limit. The formula 1/(1−r) is valid only for ∣r∣<1; at r=−1 it gives 21, which is the average of the partial sums (a hint of “Cesàro summation”, met in the problems), but not a sum.
Check yourselfEnter a number
For 1+21+41+⋯, what is the smallest n with s−sn<0.001?
Why
s−sn=rn/(1−r)=(1/2)n/(1/2)=21−n. We need 2n−1>1000, and 29=512, 210=1024, so n−1=10, n=11.
Check yourselfTrue or false?
The computation s=1+rs, hence s=1/(1−r), proves that 1+r+r2+⋯=1/(1−r) for every r=1.
Why
False. It assumes that the series has a sum s to begin with. For r=2 it would give 1+2+4+⋯=−1. Once convergence is known (for ∣r∣<1, from the formula for sn), the computation correctly identifies the sum.
Check yourselfChoose one
Which fraction is ⋅363636…?
Why
A two-digit block recurring gives A/(102−1)=36/99=4/11. The denominator is 99, one less than 10n for a block of n=2 digits; 36/999 would be ⋅036036…, and 36/100=9/25 is the terminating decimal ⋅36.
Hardy’s language
Summary
Summary
For r=1, 1+r+⋯+rn−1=1−r1−rn, so the series behaves exactly as rn does.
It converges to 1/(1−r) iff −1<r<1; diverges to +∞ iff r≥1; oscillates finitely at r=−1 and infinitely for r<−1.
The error after n terms is rn/(1−r): each new term multiplies the error by r.
Recurring and terminating decimals are exactly the rational numbers; so a decimal that does neither is irrational.