[pymvpa] SOM with custom distance metric
alan.johnson at nyu.edu
Mon Aug 13 21:05:12 UTC 2012
Will do, but I'm just wondering someone with more experience using SOMs could tell me if the issue is as simple as it seems, or if there is more subtlety to using a different metric than modifying _get_bmu()
On Aug 11, 2012, at 3:01 PM, Michael Hanke <mih at debian.org> wrote:
> On Fri, Aug 10, 2012 at 11:45:57AM -0400, Alan Johnson wrote:
>> I'd like to use the PyMVPA self-organizing map module, but with a
>> different distance metric for input vectors besides Euclidean.
>> Looking at the source code of som.py, can I subclass SimpleSOMMapper,
>> just changing _get_bmu()? On a more theoretical note, does the metric
>> need to strictly be a metric, as opposed to a dissimilarity measure?
> This code is still from the early days. Instead of subclassing it just
> for your needs, I'd like to encourage you to extent it with the
> necessary flexibility that you need and send us a patch or a pull
> request. Your effort would be most welcome!
> Michael Hanke
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
More information about the Pkg-ExpPsy-PyMVPA