BUG: Allow all appropriate links in a Family

[tncowart]

The list of links in a Family (Poisson, Gaussian, etc.) included the
"aliased" link type instead of the primary link type, when possible.
Poisson, for example, include link.log instead of link.Log.

Using the aliased class in the family link list results in suprising
behavior like this:

In [1]: import statsmodels.genmod.families as F

In [2]: import statsmodels.genmod.families.links as L

In [3]: F.Poisson(link=L.log())
Out[3]: <statsmodels.genmod.families.family.Poisson at 0x111da15b0>

In [4]: F.Poisson(link=L.Log())
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-4-47050bb1e621> in <module>
----> 1 F.Poisson(link=L.Log())

~/Documents/old/statsmodels/statsmodels/genmod/families/family.py in __init__(self, link)
    399         if link is None:
    400             link = L.log()
--> 401         super(Poisson, self).__init__(link=link, variance=Poisson.variance)
    402
    403     def _resid_dev(self, endog, mu):

~/Documents/old/statsmodels/statsmodels/genmod/families/family.py in __init__(self, link, variance)
     83             raise TypeError(warnmssg)
     84         else:
---> 85             self.link = link
     86         self.variance = variance
     87

~/Documents/old/statsmodels/statsmodels/genmod/families/family.py in _setlink(self, link)
     64             if not validlink:
     65                 errmsg = "Invalid link for family, should be in %s. (got %s)"
---> 66                 raise ValueError(errmsg % (repr(self.links), link))
     67
     68     def _getlink(self):

ValueError: Invalid link for family, should be in [<class 'statsmodels.genmod.families.links.log'>, <class 'statsmodels.genmod.families.links.identity'>, <class 'statsmodels.genmod.families.links.sqrt'>]. (got <statsmodels.genmod.families.links.Log object at 0x111e7efa0>)

L.log works but L.Log doesn't, even though L.log is an simply an "alias"
of L.Log.

By changing the family link lists to use "unaliased" classes (the parent
classes of the various aliases) both L.log and L.Log work as expected.

• closes #xxxx
• tests added / passed.
• code/documentation is well formatted.
• properly formatted commit message. See
  NumPy's guide.

Notes:

• It is essential that you add a test when making code changes. Tests are not
  needed for doc changes.
• When adding a new function, test values should usually be verified in another package (e.g., R/SAS/Stata).
• When fixing a bug, you must add a test that would produce the bug in main and
  then show that it is fixed with the new code.
• New code additions must be well formatted. Changes should pass flake8. If on Linux or OSX, you can
  verify you changes are well formatted by running

  git diff upstream/main -u -- "*.py" | flake8 --diff --isolated
  

  assuming flake8 is installed. This command is also available on Windows
  using the Windows System for Linux once flake8 is installed in the
  local Linux environment. While passing this test is not required, it is good practice and it help
  improve code quality in statsmodels.
• Docstring additions must render correctly, including escapes and LaTeX.

[pep8speaks]

Hello @tncowart! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2021-10-22 01:16:37 UTC

[josef-pkt]

unrelated random failure, missing seed

[josef-pkt]

define the list outside, then it doesn't need to be repeated

cases = [....

[josef-pkt]

Thanks, I think this looks good.
I'm running into this also every once in a while.

closes #1509

[josef-pkt]

all green and LGTM, merging

@tncowart Thank you for the PR