Olympiad Maths Prep

Track / Stage 6 / 150 of 400 #1150 of 2000

Problem 1150

National olympiad, first round
Number theory Difficulty 6.2 Prove it

4.15. (New York, 73). A finite set BRB \subset \mathbf{R} is called a basis for a set MRM \subset \mathbf{R} if each number in the set MM can be uniquely represented as a product of integer powers of numbers from the set BB. Is it true that for any finite set of positive numbers, there exists a basis?

This one wants a proof. Work it on paper, read the official solution, then mark yourself honestly — the ladder only means something if the record is true.

Official solution

4.15. We will prove that for a finite set MM of positive numbers, there exists a basis BB. We will call a subset SS of positive numbers a superbasis for MM if each number in MM can be represented as a product

α1i1αmim, where αi,,αmS,ii,,imZ \alpha_{1}^{i_{1}} \ldots \alpha_{m}^{i_{m}}, \text { where } \alpha_{i}, \ldots, \alpha_{m} \in S, i_{i}, \ldots, i_{m} \in \mathbf{Z}

For example, the set MM itself is a superbasis for MM. Among all superbases for MM, we choose the set

S0={βi;;βn} S_{0}=\left\{\beta_{i} ; \ldots ; \beta_{n}\right\}

containing the minimum number of elements. We will prove that if n2n \geqslant 2, then S0S_{0} is a basis for MM. Suppose that some element uMu \in M admits different representations as a product of integer powers of elements from S0S_{0}:

u=β1i1βnin=β1f1βnin, i.e. β1k1βnkn=1 u=\beta_{1}^{i_{1}} \ldots \beta_{n}^{i_{n}}=\beta_{1}^{f_{1}} \ldots \beta_{n}^{i_{n}}, \quad \text { i.e. } \quad \beta_{1}^{k_{1}} \ldots \beta_{n}^{k_{n}}=1

for integers kl=illlk_{l}=i_{l}-l_{l}, not all zero for l=1,,nl=1, \ldots, n. Without loss of generality, we can assume that kn0k_{n} \neq 0. Let

S1={γ1,,γn1}, where γl=βl1/kn for l=12,n1. S_{1}=\left\{\gamma_{1}, \ldots, \gamma_{n-1}\right\}, \quad \text { where } \gamma_{l}=\beta_{l}^{1 / k_{n}} \text { for } l=1_{2} \ldots, n-1 .

Then each element of the set S0S_{0} can be represented as a product of integer powers of elements from S1S_{1}:

βl=γlkn for l=1,,n1;βn=γ1k1γn1kn1 \beta_{l}=\gamma_{l}^{k_{n}} \text { for } l=1, \ldots, n-1 ; \beta_{n}=\gamma_{1}^{-k_{1}} \ldots \gamma_{n-1}^{-k n-1}

Therefore, the set S1S_{1} is a superbasis for MM, containing n1n-1 elements, which contradicts the choice of S0S_{0}. Hence, S0S_{0} is a basis for MM. If there exists a superbasis S0S_{0} for MM containing a single element β1\beta \neq 1, then S0S_{0} is also a basis for MM, since the equality βi=β\beta^{i}=\beta^{\prime} is impossible for iji \neq j. Finally, if the set S0={1}S_{0}=\{1\} is a superbasis for MM, then M={1}M=\{1\} and the set S1={2}S_{1}=\{2\} will be a basis for MM.

Source: NuminaMath-1.5, licensed Apache-2.0. Statement and solution reproduced as published; topic, difficulty and ordering added by this site.