Let S n be the set of all the numbers of the form ϵ n 3 + ϵ n − 1 3 + . . . + ϵ 2 3 + ϵ 1 3 where the number of radicals is an arbitrary non-negative integer number n and ϵ k = ± 1 for any non-negative integer k less than or equal to n , and S = ∪ n = 1 ∞ S n . Find the largest positive real number c such that ∣ x ∣ > c for all x ∈ S . Then enter ⌊ 1 0 0 0 0 0 0 c ⌋ as your answer. If c does not exist, then enter 0. Justify your answer.
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
It should be 2 1 ( 5 − 1 3 ) in the last line
We can introduce the following notation x n ( ϵ 1 , ϵ 2 , . . . , ϵ n ) = ϵ n 3 + ϵ n − 1 3 + . . . + ϵ 2 3 + ϵ 1 3 , ( ∗ ) for any non-negative integer number n , and any finite sequence ( ϵ k ) 1 ≤ k ≤ n , where ϵ k = ± 1 for any integer k satisfying 1 ≤ k ≤ n . Of course, S is the set of all possible numbers of form (*). It easy to prove by Mathematical Induction that all these number are real. We are going to use also the following notation y n = x n ( 1 , 1 , 1 , . . . , 1 ) , where the last expression contains n ones. We are going to divide the proof into three parts.
Part 1. To prove that the sequence y n has the limit r = 2 1 + 1 3 and y n < r for any non-negative integer n .
We can see that y n + 1 2 = 3 + y n , and using mathematical induction and this recursion, we can prove that the sequence ( y n ) is increasing and is bounded above by the number r . Since this sequence is increasing, upper-bounded and all its members are positive, then it has positive limit l and from the recursion (taking limits on both sides) we obtain that l satisfies the equation l 2 = 3 + l , therefore l = 2 1 ± 1 3 . When you picked the negative sign in front of the radical you get a negative value that cannot happen, therefore, l = 2 1 + 1 3 = r . Of course, it is clear that as the sequence ( y n ) is increasing and its limit is r , then y n < r for all values of n .
Part 2. To prove that ∣ x n ( ϵ 1 , . . . , ϵ n ) ∣ ≤ y n for all non-negative integer n .
This is another thing that is easy to prove by Mathematical Induction and, of course, equality is reached when ϵ k = 1 for any k .
Part 3. To prove that the largest positive number positive integer c such that ∣ x ∣ > c for all x ∈ S is 3 − r .
Let n be whatever integer number such that n ≥ 1 . If n = 1 then ∣ x 1 ( ϵ 1 ) ∣ = y 1 = 3 > c If n > 1 , then ∣ x n ( ϵ 1 , . . . , ϵ n ) ∣ = 3 + x n − 1 ( ϵ 1 , . . , ϵ n − 1 ) ≥ 3 − ∣ x n − 1 ( ϵ 1 , . . , ϵ n − 1 ) ∣ . ( ∗ ∗ ) Using Part 2 , and Part 1 , we have that ∣ x n − 1 ( ϵ 1 , . . , ϵ n − 1 ) ∣ ≤ y n − 1 < r Then 3 − ∣ x n − 1 ( ϵ 1 , . . , ϵ n − 1 ) ∣ > 3 − r , and this together with (**) implies that ∣ x n ∣ > 3 − r = c . To see that 3 − r is the largest positive number satisfying the previous inequality it is enough to see that the numbers 3 − y n belong to S and lim n → ∞ 3 − y n = 3 − r = c .
Then c = 3 − r = 3 − 2 1 + 1 3 . Then the answer to this problem is 8 3 4 9 9 9 .
Ha, snap! You beat me to it. Our solutions are slightly different so I'll leave mine up. Can we get all numbers in the interval?
Log in to reply
Thank you, Chris! What do you mean by the numbers in the interval?
Log in to reply
Sorry, I wasn't very clear there. I mean is the set S precisely the interval [ c , r ] (using your notation)? If we choose any value u in that interval, is there a corresponding element of S ? Is there an algorithm to find that element?
By the way, since S is the union of all (infinitely many) S n , isn't it correct to say c is in S ?
Log in to reply
@Chris Lewis – I mean any solution corresponds to a root in an integer coefficient polynomial (repeatedly square are subtract 3) so numbers like pi are impossible(transcendental)
Log in to reply
@Razzi Masroor – You are right!
Log in to reply
@Arturo Presa – Yes! You are right. Thanks!
What I hadn't noticed was that we have a countable union of countable sets, so there are certainly real numbers that can't be represented.
Problem Loading...
Note Loading...
Set Loading...
Let's look at the related problem of finding y = sup S . It's easy to see that, for any given n , the largest possible number of this type is found when all ϵ k are + 1 , that is y n = 3 + 3 + 3 + ⋯ + 3 (where the expression contains n 3 s). As ever, we have to be cautious around convergence. To make this sequence easier to handle, note that y 1 = 3 and y n + 1 = 3 + y n
y n is always positive. Define f ( t ) = 3 + t ; then f ( t ) has a fixed point at t c = 2 1 + 2 1 3 . Since ∣ f ′ ( t c ) ∣ < 1 , this is an attracting fixed point; in other words, whatever positive start value is chosen for y 0 , the limit as n → ∞ of y n will always be t c .
So y = sup S = t c .
To solve the actual problem, note that the smallest (positive) number we can possibly make for a given n has the form x n = 3 − 3 + 3 + ⋯ + 3 (again with n 3 s in the expression). In the limit, this is just x = 3 − t c which works out to be x = 2 1 ( 5 − 1 3 ) = 0 . 8 3 4 9 9 9 6 …