[SciPy-Dev] Deprecate stats.glm?

[josef.pktd at gmail.com]

On Thu, Jun 3, 2010 at 11:32 AM, Nathaniel Smith <njs at pobox.com> wrote:
> On Thu, Jun 3, 2010 at 6:38 AM,  <josef.pktd at gmail.com> wrote:
>> On Thu, Jun 3, 2010 at 8:50 AM, Warren Weckesser
>> <warren.weckesser at enthought.com> wrote:
>>> stats.glm looks like it was started and then abandoned without being
>>> finished.  It was last touched in November 2007.  Should this function
>>> be deprecated so it can eventually be removed?
>>
>> My thoughts when I looked at it was roughly:
>> leave it alone since it's working, but don't "advertise" it because we
>> should get a better replacement.
>> similar to linregress the more general version will be available when
>> scipy.stats gets the full OLS model.
>
> Wait, what does 'glm' have to do with OLS (or t-tests) anyway? Surely
> if anything it *should* be a function that fits, you know, GLMs
> (generalized linear models)?
>
> I guess this is a vote for removing it, because GLMs are one of the
> fundamental stats models that people will look for, and having some
> weird, broken, other thing in the obvious place is just confusing and
> looks really bad.

That was my initial impression a long time ago.

GLM as in general linear model not generalized. (It's the worst
conflicting acronym in stats).

The function actually estimates a GLM, it construct a binary dummy
variable from the label data to get the design matrix, estimates it
with OLS, calculates the t-statistic and the corresponding p-value.

But then it becomes like the ttest_ind because it only returns the
t-statistic and the corresponding p-value.

I don't remember seeing any previous comments about it on the mailing
list, but it would be a prime candidate for "finishing" it. (except
finishing it requires a full module on it's own.)

The discussion what glm (general linear model) has to do with ols
fills already many pages on the pystatsmodels mailing list.  (GLM in
statsmodels is generalized linear model)

Josef


>
> -- Nathaniel
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev
>