There are some misunderstandings to expand continued fractions.
![]()
![]()
only for around q=1 and|a|<|b|. This does not hold all over the domain of |q|<1.
However, the following relation
![]()
![]()
only for |q|<1 and|a|<|b|.
Here, we just apply the gamma-shaped multipliers of q, q2,q3,. . . in order.
Thus, we cannot say that
![]()
![]()
only for |q|<1 and|a|<|b|. That is one big problem.
Bauer Muir transformation
![]()
![]()
for
/=0 and k=1,2,3,c
is not an identity but an approximate that is going to be the same on average.
Some of modified Rogers-Ramanujanfs and other continued fractions are identities.
Some of them are just approximations around a certain point or even based on some
wrong specification. They forge to explain or verify some wrong approximations.