After the initial release of scikits.statsmodels, we started to have
email discussions about the design, implementation and extension of
statsmodels, as well as about datahandling and how to use statsmodels
for different types of data. Since this discussion seemed to
specialized to fill up scipy-dev with it, we decided to start a
dedicated google group.
The overall objective is to make python easier to use for statistical
and econometric analysis, and the discussion group is open to and
welcomes the discussion of related packages. The first example is
pandas, a package for handling panel data that uses statsmodels for
My $.02. Mostly right now, we are only discussing design issues for
statsmodels and not statistical issues (though of course this will
come up while we're extending the models). Perhaps statistical issues
related to the existing statsmodels code and scipy are best discussed
on scipy-user, but questions about development of and extending
statistics and models in statsmodels would be appropriately discussed
there? We just didn't want to post all day to scipy-user about the
use of decorators in the model results classes for our scikit for
I wouldn't be against discussing statistical issues there, but they're
probably best kept on the scipy list to take advantage of everyone's
Anything statistics/econometrics/python related is welcome, the topics will
depend a lot on the participants.
However, I think discussion related to scipy.stats directly, bugs, problems,
improvements, usage question, should stay on the scipy lists, since it
makes it easier to find scipy.stats questions and answers in the archives
and it concerns the entire scipy community.
> On Mon, Sep 7, 2009 at 2:27 PM, Tim
> Michelsen<[hidden email]> wrote:
>>> Anyone interested, is welcome to join us at
>>> http://groups.google.ca/group/pystatsmodels/topics?hl=en >> Cool.
>> Will this be restricted to statsmodel only or can general statistical
>> matters related to scipy discussed there. too?
> Anything statistics/econometrics/python related is welcome, the topics will
> depend a lot on the participants.
> However, I think discussion related to scipy.stats directly, bugs, problems,
> improvements, usage question, should stay on the scipy lists, since it
> makes it easier to find scipy.stats questions and answers in the archives
> and it concerns the entire scipy community.
>> Would you please add the list to Gmane?
> I have to figure out how to do this.