Re: [Gretl-devel] convolve function into extra?
On Thu, 21 Mar 2019, Sven Schreiber wrote:
Am 21.03.2019 um 07:28 schrieb Riccardo (Jack) Lucchetti:
> On Wed, 20 Mar 2019, Allin Cottrell wrote:
>
>> Right, any coefficients beyond the degree of either of the two
>> polynomials are taken to be zero. Perhaps that should be explicit. Jack?
>
> I don't think we have to be explicit on that; after all, that's the very
> definition of order of a polynomial. IMO the text in its present form is
> perfectly OK.
I agree the text is OK (present = git), but just to be clear while we're
at it: k-n will be negative for low k, and when for example p_{-1} is
not defined, it's not always totally obvious that it will also be
understood as zero [...]
Where does k-n arise? The r_k formula runs the p-index from 0 to k and
the q-index from k to 0, for each k in [0, m+n], so I don't see how
iteration could throw up a negative index.
Allin