[Octave-patch-tracker] [patch #8517] Legendre-Papoulis filter design (for signal package)

[anonymous]

Follow-up Comment #2, patch #8517 (project octave):

> Would you mind looking at some coding style fixups and adding unit tests?

Sure, I don't mind. Well, at the time I just blatantly copied the 'butter'
function and change it to use 'legendap', as well as change its doc. Moreover
I worked with a pretty old GNU Octave. So merely re-copying the 'butter'
function should fix most coding style issues, and I'll fix the rest.

A bit unrelated, but I realized I forgot to add the warning about numerical
accuracy, should I mention it in 'legendap' and refer to it in 'legendf'?

> Would you agree to rename these functions to "legendreap" and "legendref"? I
think it would be clearer to distance these functions from the plotting
function "legend".

I liked 'legendap' and 'legendf' as they are short names and reads somewhat
well. 'legendreap' looks funny. On the other hand, this is how other filter
design functions are named:

* Butterworth: butter, buttap
* Chebyshev type 1,2: cheby[12], cheb[12]ap
* Elliptic: ellip, ellipap

In the same spirit, it could be 'legenp' and 'legenpap' (p for Papoulis). What
do you think?

I realize also that there should be a 'legenpord' to match the
'buttord','ellipord',etc. functions. This will be a candidate of a future
patch however.

    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/patch/?8517>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/