Smarter way to calculate middle-right terms of continued fraction partial sums
- To: mathgroup at smc.vnet.net
- Subject: [mg70024] Smarter way to calculate middle-right terms of continued fraction partial sums
- From: "Diana" <diana.mecum at gmail.com>
- Date: Sat, 30 Sep 2006 05:13:05 -0400 (EDT)
Math folks
I am trying to calculate the iterative terms of a continued fraction.
It is a limit of partial sums of continued fractions defined as
follows:
x[n_]:= Sum as j goes from 1 to n of (z^q^j)/d[j], where d is a
function of j.
When n = 1, z = 1, and q = 2,
x[1]:= 1/d[1]
I then
Solve[x[1] == 1/a1, a1]
So, x[1] = {0, a1}, a two-term continued fraction.
The second term looks like:
x[2]:= 1/d[1] + 1/d[2]
I then
Solve[x[2] == 1/(a1 + (1/-a2 + (1/-a1))), a2]
And get a partial sum which looks like
x[2] = {0, a1, -a2, -a1}, a four-term continued fraction.
x[3]:= 1/d[1] + 1/d[2] + 1/d[3]
Solve[x[3] == 1/(a1 + (1/-a2 + (1/-a1 + (1/-a3 + (1/a1 + (1/a2 +
(1/-a1))))))), a3]
x[3] = {0, a1, -a2, -a1, -a3, a1, a2, -a1}, an eight-term continued
fraction.
So, each successive x[n] term is two times as long as the one before
it, and I am always solving for the middle right term. After I have
found the x[3] term, for example, I can plug the solved value of a3
into the Solve equation and solve for the x[4] term.
The terms eventually look like the following:
{0, a1, -a2, -a1, -a3, a1, a2, -a1, -a4, a1, -a2, -a1, a3, a1, a2, -a1,
-a5, a1, -a2, -a1,
-a3 a1, a2, -a1, a4, a1, -a2, -a1, a3, a1, a2, -a1, a6, ...}
Note that the signs of a1 alternate between positive and positive.
Likewise, all the variables alternate in sign.
I have been coding the Solve equation as fractions of 1 over a1 + 1
over -a2 + 1 over, etc.
By the fifth term, I have 63 fractions to code. I am sure there is a
smarter way to do this, so I am posting this question to the news
group.
Can someone help?
Diana