[Biopython-dev] Numpy conversion

[Michiel de Hoon]

Bio.kNN is the only module that imports Bio.distance. Bio.distance is written in Python, but it also imports a C version of Bio.distance if it is available. From the comments in the code, I gather that the purpose of the C-version is to get fast distance calculations without using Numeric / NumPy. However, Bio.kNN itself uses Numeric / NumPy, which defeats the purpose of the C-version of Bio.distance.

I would therefore like to propose to add a NumPy-aware version of the code in Bio.distance to Bio.kNN, and to deprecate Bio.distance.

Any objections?

--Michiel.

--- On Thu, 9/18/08, Michiel de Hoon <mjldehoon at yahoo.com> wrote:

> From: Michiel de Hoon <mjldehoon at yahoo.com>
> Subject: Re: [Biopython-dev] Numpy conversion
> To: "Peter" <biopython at maubp.freeserve.co.uk>
> Cc: biopython-dev at biopython.org
> Date: Thursday, September 18, 2008, 10:10 AM
> > I've not used it myself, but it sounds handy. 
> Michiel,
> > does this overlap at all with your clustering module?
> 
> No, it doesn't. Bio.Cluster contains unsupervised
> clustering methods only. The k-nearest neighbors in Bio.kNN
> is a supervised learning method.
> 
> --Michiel.
> 
> --- On Wed, 9/17/08, Peter
> <biopython at maubp.freeserve.co.uk> wrote:
> 
> > From: Peter <biopython at maubp.freeserve.co.uk>
> > Subject: Re: [Biopython-dev] Numpy conversion
> > To: mjldehoon at yahoo.com
> > Cc: biopython-dev at biopython.org
> > Date: Wednesday, September 17, 2008, 10:29 AM
> > On Wed, Sep 17, 2008 at 3:13 PM, Michiel de Hoon
> > <mjldehoon at yahoo.com> wrote:
> > > Hi everybody,
> > >
> > > I am now looking at the pure-python modules that
> make
> > use of Numerical Python / NumPy.
> > > Bio.kNN is one of them; this also happens to be
> the
> > only module that imports Bio.distance,
> > > which also depends on NumPy.
> > >
> > > What I am not sure about is the usage of Bio.kNN.
> A
> > quick google search didn't reveal much,
> > > suggesting that it is not widely used. Bio.kNN
> > currently is not documented in the tutorial, but
> > > the code itself is reasonably well documented.
> > >
> > > How do you guys feel about this module? Should we
> keep
> > it?
> > >
> > 
> > I've not used it myself, but it sounds handy. 
> Michiel,
> > does this
> > overlap at all with your clustering module?
> > 
> > Peter
> 
> 
>       
> _______________________________________________
> Biopython-dev mailing list
> Biopython-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biopython-dev