Thursday, 24 August 2017

exponentiation - Proof that irrational exponents exist, and are unique?



Can someone provide a proof that for any given irrational number, b, exponentiation by that number defined as a limit of rational powers always converges, and that if we choose a particular base, a, that the no two b's yield the same limit. (I suppose in the case where a is not 0 or 1?)


Answer



We can restrict to the case a>1 (noting that $0Note that pq<rs implies $psxax is strictly increasing when looking only at rational x. If we take the convergence for granted, this implies that xax is injective as a function defined on all of R: If bR and b<pqQ, then almost all fractions in a sequence converging to b are <pq, hence their powers are $

Now for the convergence: For rational exponents, the usual power laws hold, and as we have just seen order is also respected (for bases a>1). Hence whenever |pqrs|<1N, we have $$\left|a^{\frac pq}-a^{\frac rs}\right|=a^{\frac rs}\cdot \left|a^{\frac pq-\frac rs}-1\right|
We may assume wlog that rs<b+1=:M so that $$\left|a^{\frac pq}-a^{\frac rs}\right|As Na1 and N1/a1, we conclude that any sequence {pkqk}kN of fractions converging to b produces a Cauchy sequence {apkqk}kN.


No comments:

Post a Comment

real analysis - How to find limhrightarrow0fracsin(ha)h

How to find lim without lhopital rule? I know when I use lhopital I easy get $$ \lim_{h\rightarrow 0}...