[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] serial_to_parallel and fft_v{f,c}c

From: Eric Blossom
Subject: Re: [Discuss-gnuradio] serial_to_parallel and fft_v{f,c}c
Date: Thu, 16 Mar 2006 15:15:58 -0800
User-agent: Mutt/1.5.9i

On Thu, Mar 16, 2006 at 11:44:58PM +0100, Patrick Strasser wrote:
> Eric Blossom wrote:
> >On Thu, Mar 16, 2006 at 09:45:05PM +0100, Patrick Strasser wrote:
> Windowing is clear to me, but I wasn't able to read the python datatype 
> or gnuradio window type from the C++ .h file. I suppose it's something 
> like gr.firdes.WIN_RECTANGULAR or a vector/list of window coefficients, 
> but I could not figure out exactly until now. Any hints? ;-)

Looking at the header, this is effectively the public constructor:

gr_make_fft_vfc (int fft_size, bool forward, const std::vector<float>);

  (<foo>_sptr are typedefs for "boost smart pointers".  These do
  automatic reference counting in C++ and are used behind the scenes
  to keep track of storage.  We ref count in C++ with the smart
  pointers.  Python ref counts using it's own techniques.  This
  somewhat twisted combination works.)

The constructor takes three args: an int, a bool and a vector of floats.

When mapped to python these are:

  an int: e.g., 512
  a bool: e.g., True
  a sequence (list or tuple) of floats.  E.g., (2.3, 2.4, 8.5)

The window needs to be the same length as the fft.

There are a lot of functions defined in gnuradio.window that compute
windows.  gnuradio-core/src/python/gnuradio/window.py

You're absolutely right, we need a comprehensive merged set of documentation ;)

> That's my main resource, giving primary hints and leading to the source 
> files...
> >The output should be python-centric docs in epydoc output format.
> >See http://epydoc.sf.net and http://www.stack.nl/~dimitri/doxygen
> >
> >I've got a patch somewhere in my email that's supposed to do this.
> >If you're interested in testing it out, I'll dig it out and you can
> >try it.  If it works, we'll apply it.
> I'd really like to try this.

I'll look for it in the next day or so.

> You know, I'm hindered in my wiki-I-want-to-add-more-doc behaviour. Good 
> docs are essential for beginnerns like me, and people used to the system 
> most likely don't see what's _not_ obvious...

Yes.  I'm glad you're asking all these questions.  Now we just need to
record the answers where the next person can find them ;)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]