Help: filter built-in function and initial condition

[Nicola De Quattro]

Hi all,
I'm a Octave user from Italy. I've a doubt using function filter: in
my university studies, initial conditions for a filter are values of
output at negative steps. For example, if N is the order of the
filter, I know that initial condition (ic) is a vector such as:

ic(1:N)=y(-N,...,-1)

So applying the canonical formula to calculate filter output as
reported (I hope you can see it, but you can have it viewing filter
function help):

                N                             M
 y(n) = - SUM c(k) y(n-k) + SUM d(k) x(n-k),   for 1<=n<=length(x)
               k=1                        k=0


I think this initial conditions will only be value of y(n-k) for n-k<0
(for this value x(n-k) is assumed to be 0, causal signal in input).

Is it correct? I'm not sure 'cause there are some results different
from those I expected. I'm not good with C++ (I refer to code in
Octave Function reference) so I can't verify...

Thank you very much

Nik