Re: closed form for convolution of a^n with Fibonacci numbers
- From: David C. Ullrich <ullrich@xxxxxxxxxxxxxxxx>
- Date: Sun, 17 Jul 2005 08:44:17 -0500
On 16 Jul 2005 18:17:53 -0700, "hawkmoon269" <rson@xxxxxxxxxx> wrote:
>Thanks for the response. I'm sorry, though...I know the closed form
>and the formula for the sum of a geometric series, but I don't
>understand how this could be done...because I have to compute either
>the n-kth Fibonacci or a^(n-k)...i.e., I don't know how to deal with
>k...
Say f_n is the n-th Fibonacci number. Then that closed form
says
f_n = a b^n + c d^n
where a, b, c, and d are constants involving all sort of square roots,
right?
Now you wanted the convolution of this sequence and the sequence
e_n, where
e_n = g^n
for g some integer.
The convolution would be
h_n = sum_{k=0}^n e_k f_{n-k}.
If you plug in the two formulas for f_n and e_n you see that
h_n = a b^n sum (g/b)^k + c d^n sum (g/d)^k,
where both sums are from k = 0 to k = n.
Those sums are finite geometric series, so you can write
down a closed form for them and you're done.
************************
David C. Ullrich
.
- Follow-Ups:
- Re: closed form for convolution of a^n with Fibonacci numbers
- From: hawkmoon269
- Re: closed form for convolution of a^n with Fibonacci numbers
- References:
- closed form for convolution of a^n with Fibonacci numbers
- From: hawkmoon269
- Re: closed form for convolution of a^n with Fibonacci numbers
- From: David C . Ullrich
- Re: closed form for convolution of a^n with Fibonacci numbers
- From: hawkmoon269
- closed form for convolution of a^n with Fibonacci numbers
- Prev by Date: Re: Relative Cardinality
- Next by Date: Re: I am confused by the diagonal argument
- Previous by thread: Re: closed form for convolution of a^n with Fibonacci numbers
- Next by thread: Re: closed form for convolution of a^n with Fibonacci numbers
- Index(es):
Relevant Pages
|