Non-uniqueness of integers with non-integer bases

In this post I’ll be exploring non-integer bases and the “natural numbers” they represent. I say “natural numbers” because the resulting sums won’t necessarily be integers if the base is not an integer. I’ll use N={0,1,2,} and for kN I’ll use Nk={k,k+1,}. I’ll denote the base under consideration by β and throughout I’ll be assuming β>1. Given a base I’ll have some digit set Dβ={0,1,,dβ} with dβ being the largest digit. It will always be the case that 0,1Dβ. When βN2 it’ll always be the case that dβ=β1.

The term “base-β representation” of a number x[0,) will denote a finite sequence (a0,a1,,an)Dβn with an>0 such that
x=k=0nakβk.

Result 1: when βN2 then each element of N has a unique base-β representation.

Pf: let x[0,) and suppose x can be represented with either digit sequence a or b, i.e.
x=k=0nakβk=k=0mbkβk
First I’ll show that n=m, i.e. a and b have the same length.

Suppose n>m. Since an>0 (by definition) the smallest that k=0nakβk can be is if an=1 and ak=0 for k<n. This means k=0nakβk=βn. On the other hand, the largest that k=0mbkβk can be is if bk=dβ for all k, which means k=0mbkβk=(β1)k=0mβk=(β1)1βm+11β=βm+11. m+1n so the largest that βm+1 could be is βn, but because of the 1 term there’s no way to get k=0mbkβk large enough to catch up to k=0nakβk if a is allowed even one extra digit. The exact same argument shows m>n also leads to a contradiction, therefore n=m.

Now I’ll suppose an>bn, where n=m means these are each the largest digits. By a similar argument, the largest that k=0nβk can be with bn fixed at less than an is
(β1)k=0n1βk+bnβn=βn1+bnβn=(bn+1)βn1.
But the smallest that k=0nakβk can be is anβn and anbn+1 so again the 1 term makes it so that there’s no way that k=0nβk can catch up if an>bn. This means that an=bn, and applying this downward shows that a=b, so x has a unique base-β representation.

◻

I’ll now show that this does not apply to non-integer bases.

Result 2: a base exists such that the same number has two different representations.

Pf: let φ=1+52 (so φ is the golden ratio) and consider 100φ vs 11φ. I know that φ is the positive root of the equation x2x1=0, or equivalently it satisfies x2=x+1, so
100φ=φ2=φ+1=11φ
so these two distinct digit sequences lead to the same number.

◻

Now I want to see how unique this property is. If Dβ={0,1} then φ is the unique base that makes 100β=11β, but what about for longer digit sequences?

 

Finding bases for non-unique integers

I’ll now look at the following problem: for any nN2 is there a base βn>1 such that
100βn+1=ddβn
i.e. the number with n+1 digits formed by a 1 followed by n zeros (the smallest number with n+1 digits) equals the number formed by n of the largest digit (the largest number with n digits). I’m using d=dβ to make the notation clearer. I’ll require 1d<β but I’ll think of it as otherwise unconstrained. For example, it’d be legal to choose β=π but restrict myself to just the digits {0,1}. This turns out to be a problem of polynomials: assuming β1, I’m looking for βn=k=0n1dβk=d1βn1ββnβn+1=ddβnβn+1(d+1)βn+d=0. Thus I’ll consider the polynomial pn,d(x)=xn+1(d+1)xn+d and I’ll look for roots greater than d. This restriction encodes the fact that valid bases are all greater than the maximal digit d, which itself is at least 1, so that also handles the β>1 requirement. I want roots because this means the two digit sequences under consideration are equal.

First I’ll note that
pn,d(d+1)=(d+1)n+1(d+1)n+1+d=d>0
for any n and d, and
pn,d(x)=(n+1)xnn(d+1)xn1=xn1[(n+1)xn(d+1)].
For xd+1 pn,d is positive which means pn,d is increasing and therefore has no roots in [d+1,). This makes sense because when the base is that much larger than the largest digit then the smaller digit sequences can’t get big enough to equal a longer digit sequence.

Next, for x=d I have
pn,d(d)=dn+1dn+1dn+d=d(1dn1)
so if d2 then pn,d<0 while if d=1 I have pn,1(1)=0. By the Intermediate Value Theorem (IVT) this establishes that pn,d has a root in (d,d+1) for every nN2 as long as d2.

I’ll prove the following lemma which will finish the d=1 case.

Lemma: Let f:RR be differentiable. If for some aR I have f(a)=0 and f(a)<0 then there is some h>0 such that f(a+h)<0.

Pf: by the definition of the derivative I have
limh0f(a+h)f(a)h=limh0f(a+h)h=L
for some L<0. Formally, this means for all ε>0 there is some δ>0 such that 0<|h|<δ implies
|f(a+h)hL|<ε. This also means that for any ε>0 there is a δ>0 such that 0<h<δ implies this same result (this just means that the two-sided limit existing and equaling L implies the same for both one sided limits). h>0 so the sign of f(a+h)h is just the sign of f(a+h). I can make f(a+h)h arbitrarily close to L, which is negative, so it must be that for δ>0 sufficiently small, f(a+h) is negative on (0,δ).

◻

This result establishes that even though pn,1(1)=0, I have pn,1(1)<0 so pn,1(1+h)<0 for a sufficiently small h>0. Then applying the IVT again tells me there is a root in (1,2).

I’ve determined that no matter the maximal digit I’ll allowed to use, and the digit string in question (so long as it is at least length 2).

I can also show that this base is unique, given d and n. I know pn,d<0 at d (except for when d=2, but pn,1 becomes negative immediately after). Because pn,d=xn1[(n+1)xn(d+1)] I know there’s a critical point at x^=nn+1(d+1) and that pn,d decreases to this point so it’s a minimum. This means there can’t be any roots in (d,x^]. I already know there is one root in [x^,d+1) but because p>0 for x>x^ this means that p is monotonically increasing past x^ so that one root is the only one.

A few examples: first, if n=2 then I want the root to
p2,1(x)=x32x2+1=(x1)(x2x1)
so I do still have φ as the unique positive root larger than d=1.

Next, suppose I take d=7 to be my largest allowed digit. I know for nN2 there is a base β that makes
100β=77β
and I can find it explicitly: this equation tells me β27β7=0 so
β=7+772
where I’ve dropped the negative root as it’s not meaningful here.

If instead I try
1000β=777β

I need to solve
β37β27β7=0
which has as its unique real root
β=13[7+65842513+14(47+351)3]7.986
(I just used Wolfram|Alpha for this).

Leave a Reply

Your email address will not be published. Required fields are marked *