Product of sigma(k)/phi(k)

[Paul D Hanna]

To be specific, the continued fraction is:
 
[1; x - 1, x + 2, x, x, x - 2, x, x + 2, x, x - 2, x + 2, x, 
x - 2, x, x, x + 2, x, x - 2, x + 2, x, x, x - 2, x, x + 2, 
x - 2, x, x + 2, x, x - 2, x, x, x + 2, x, x - 2, x + 2, x, 
x, x - 2, x, x + 2, x, x - 2, x + 2, x, x - 2, x, x, x + 2, 
x - 2, x, x + 2, x, x, x - 2, x, x + 2, x - 2, x, x + 2, x, ...]
  
Is this already in the OEIS?


-- Ralf Stephan <ralf at ark.in-berlin.de> wrote:
> (Does sum {1/x^(2^i)} have a known formula?)

It has continued fraction coefficients satisfying a
bifurcating (d&c) recurrence, for integer x. I conjecture it
is possible to write the cont.frac. in the unknown using
such a recurrence, if Shallit and/or van der Poorten have 
not already done so.

ralf