On Tue, May 28, 2013 at 10:45 PM, Salman Javaid <address@hidden> wrote:
Should it just be part of the signal processing package?
Well, yes of course. I was looking at it from Matlab's perspective.
It's also very important that you can affirm that you have not
consulted any code from The Mathworks while writing this package.
I can affirm to that. In particular, Matlab's 2011 version doesn't implement
Bayesian Shrink, and NeighShrink that I plan to implement. SureShrink is
implemented though in Matlab.
Best Regards,
Salman Javaid
On Tue, May 28, 2013 at 1:15 PM, Jordi Gutiérrez Hermoso
<address@hidden> wrote:
On 28 May 2013 16:05, Salman Javaid <address@hidden> wrote:
I will like to develop a Wavelet Denoising module for Octave. Is
this something that Octave team will be interested in?
Sure.
Should it just be part of the signal processing package?
By the way, you should note that TMW claims trademarks on a bunch of
things that end with "toolbox":
http://www.mathworks.com/company/aboutus/policies_statements/trademarks.html
It is for this reason that we generally avoid the word "toolbox" when
talking about Octave packages.
It's also very important that you can affirm that you have not
consulted any code from The Mathworks while writing this package.
- Jordi G. H.
This can be cool, sure we are interested.
If you think the package has enough structure to live on its own. I
would recommend you package it into a "wavelet" package.
You can also add it as a subfolder to the signal package, no problem
with that as well.
The important thing is not whether you implement a function that is
already in Matlab, but rather that your implementation is independent
of mathworks' one. You can implement the same algorithm, but you
shouldn't copy their implementation (nor they help-string, etc etc..).
If you think you have already somethign to you you can put it in Agora
so we can take a look at it.
http://agora.octave.org/
Do not forget to send us the link.