Use iteration to guess an explicit formula

  • Thread starter Thread starter Arew
  • Start date Start date
  • Tags Tags
    Explicit Formula
AI Thread Summary
The discussion revolves around finding an explicit formula for the recurrence relation u_k = u_{k−2} * u_{k−1}, with initial conditions u_0 = u_1 = 2. The participant calculates the first few terms and conjectures that the general form is 2^{F_k}, where F_k represents Fibonacci numbers. There is a suggestion to use induction for proving the conjectured formula. Additionally, there are technical discussions about formatting mathematical expressions, indicating some confusion about using MathJax for rendering. The participant expresses satisfaction with their progress but seeks confirmation on the correctness of their approach.
Arew
Messages
7
Reaction score
0

Homework Statement



Use iteration to guess an explicit formula for u_k = u_{k−2} * u_{k−1}, for all integers k ≥ 2, u_0 = u_1 = 2 and prove it .

Homework Equations



Hint: Express the answer using the Fibonacci sequence.

The Attempt at a Solution



u_k = u_{k−2} * u_{k−1} and u_0 = u_1 = 2, so

u_2 = 2^2

u_3 = 2^3

u_4 = 2^5

u_5 = 2^8

u_6 = 2^13

u_7 = 2^21

Then (I think) in general we have ,

2^{F_k} = 2^{F_k-1 + F_k-2}

Not finished yet... But does it make sense so far?

BTW, I tried putting $ signs around expressions, but mathjax(?) doesn't seem to render in preview. [/B]
 
Physics news on Phys.org
Arew said:
But does it make sense so far?
Yes.
Arew said:
BTW, I tried putting $ signs around expressions, but mathjax(?) doesn't seem to render in preview.
That was dollar signs in pairs, right? $$latex$$. Not sure if mathjax is same as LaTeX.
 
haruspex said:
Yes.

That was dollar signs in pairs, right? $$latex$$. Not sure if mathjax is same as LaTeX.

Oh, I never knew you had to put the dollar signs in pairs. Thanks.$$2^{F_k} = 2^{F_{k-1} + F_{k-2}}$$ then $$2^{F_n} = 2^ {\frac{\frac {1 + \sqrt 5}{2} - \frac {1 + \sqrt 5}{2}}{\sqrt 5}}$$ which implies $$\log_22^{F_n} = \log_22^ {\frac{\frac {1 + \sqrt 5}{2} - \frac {1 + \sqrt 5}{2}}{\sqrt 5}}$$ meaning $$F_n = {\frac{\frac {1 + \sqrt 5}{2} - \frac {1 + \sqrt 5}{2}}{\sqrt 5}}$$

Can I claim this to be the solution to the given recurrence relation?
 
Arew said:

Homework Statement



Use iteration to guess an explicit formula for u_k = u_{k−2} * u_{k−1}, for all integers k ≥ 2, u_0 = u_1 = 2 and prove it .

Homework Equations



Hint: Express the answer using the Fibonacci sequence.

The Attempt at a Solution



u_k = u_{k−2} * u_{k−1} and u_0 = u_1 = 2, so

u_2 = 2^2

u_3 = 2^3

u_4 = 2^5

u_5 = 2^8

u_6 = 2^13

u_7 = 2^21

Then (I think) in general we have ,

2^{F_k} = 2^{F_k-1 + F_k-2}

Not finished yet... But does it make sense so far?

BTW, I tried putting $ signs around expressions, but mathjax(?) doesn't seem to render in preview. [/B]
Suggest you google Fibonacci series. I think you might have already solved this one as much as you can in post #1. (You still need to prove that your guess is correct.)
 
Last edited:
Charles Link said:
Suggest you google Fibonacci series. I think you might have already solved this one as much as you can in post #1. (You still need to prove that your guess is correct.)

I was unsure. Usually when we find an explicit formula for a recurrence, we get rid of indices in the recurrence relation so that we do at most one substitution into the formula. But if you say I am done - I am happy to take a break :) As for the proof, I think induction should take care of it. Thanks.
 
Back
Top