180 IV Continued Fractions and Their Uses
with matrix
T′=
(
α′ β′
γ′ δ′
)
,
then, as is easily verified, the matrix
T′′=
(
α′′ β′′
γ′′ δ′′
)
of the composite transformation
ξ=(α′′ξ′′+β′′)/(γ′′ξ′′+δ′′)
is given by the matrix productT′′=TT′.
It follows that, if we set
Ak=
(
ak 1
10
)
,
then the matrix of the linear fractional transformation which expressesξin terms of
ξn+ 1 is
Tn=A 0 ···An.
It is readily verified by induction that
Tn=
(
pn pn− 1
qn qn− 1
)
,
i.e.,
ξ=(pnξn+ 1 +pn− 1 )/(qnξn+ 1 +qn− 1 ),
where the elementspn,qnsatisfy the recurrence relations
pn=anpn− 1 +pn− 2 , qn=anqn− 1 +qn− 2 (n≥ 0 ), (1)
with the conventional starting values
p− 2 = 0 , p− 1 = 1 , resp.q− 2 = 1 , q− 1 = 0. (2)
In particular,
p 0 =a 0 , p 1 =a 1 a 0 + 1 , q 0 = 1 , q 1 =a 1.
Since detAk=−1, by taking determinants we obtain
pnqn− 1 −pn− 1 qn=(− 1 )n+^1 (n≥ 0 ). (3)
By (1) also,