MAINT: update numpydoc to v1.1.0

[rgommers]

The main issue was annoying to track down. These are new warnings that showed up due to a change in numpydoc:

/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.clear:: WARNING: py:class reference target not found: None.  Remove all items from D.                                                                                                 
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.copy:: WARNING: py:class reference target not found: a shallow copy of D                                                                                                              
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.items:: WARNING: py:class reference target not found: a set-like object providing a view on D's items                                                                                 
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.keys:: WARNING: py:class reference target not found: a set-like object providing a view on D's keys                                                                                   
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.pop:: WARNING: py:class reference target not found: v, remove specified key and return the corresponding value.                                                                       
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.popitem:: WARNING: py:class reference target not found: (k, v), remove and return some (key, value) pair as a                                                                         
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.update:: WARNING: py:class reference target not found: None.  Update D from dict/iterable E and F.                                                                                    
/home/rgommers/code/scipy/scipy/optimize/__init__.py:docstring of scipy.optimize.OptimizeResult.values:: WARNING: py:class reference target not found: an object providing a view on D's values                                                                                       
/home/rgommers/code/scipy/scipy/sparse/__init__.py:docstring of scipy.sparse.dok_matrix.clear:: WARNING: py:class reference target not found: None.  Remove all items from D.                                                                                                         
/home/rgommers/code/scipy/scipy/sparse/__init__.py:docstring of scipy.sparse.dok_matrix.items:: WARNING: py:class reference target not found: a set-like object providing a view on D's items                                                                                         
/home/rgommers/code/scipy/scipy/sparse/__init__.py:docstring of scipy.sparse.dok_matrix.keys:: WARNING: py:class reference target not found: a set-like object providing a view on D's keys                                                                                           
/home/rgommers/code/scipy/scipy/sparse/__init__.py:docstring of scipy.sparse.dok_matrix.pop:: WARNING: py:class reference target not found: v, remove specified key and return the corresponding value.                                                                               
/home/rgommers/code/scipy/scipy/sparse/__init__.py:docstring of scipy.sparse.dok_matrix.popitem:: WARNING: py:class reference target not found: (k, v), remove and return some (key, value) pair as a                                                                                 
/home/rgommers/code/scipy/scipy/sparse/dok.py:docstring of scipy.sparse.dok.dok_matrix.update:: WARNING: py:class reference target not found: None.  Update D from dict/iterable E and F.                                                                                             
/home/rgommers/code/scipy/scipy/sparse/__init__.py:docstring of scipy.sparse.dok_matrix.values:: WARNING: py:class reference target not found: an object providing a view on D's values

The root cause is that dict methods don't have proper html docs that can be found via intersphinx (see https://bugs.python.org/issue11975). That is fixed by ignoring reference issues for builtins in Sphinx, see sphinx-doc/sphinx#7254.

It's not clear to me if the warnings appearing now is a regression in numpydoc, or a correct fix that made the issue show up. The two candidate changes in numpydoc that touched this behaviour are:

• BUG: Connect to earlier event numpy/numpydoc#269 (@larsoner)
• MAINT: Suggestions from reviewing test suite numpy/numpydoc#271 (@rossbar)
  Maybe this means something to one of you?

I couldn't upgrade to numpydoc latest master due to another issue, but that's for another time.

[rgommers]

Here's the change in the page generated for one of those methods:

Compare with: http://scipy.github.io/devdocs/generated/scipy.optimize.OptimizeResult.keys.html

EDIT: which is because of:

>>> dict.keys.__doc__
"D.keys() -> a set-like object providing a view on D's keys"

This is with Sphinx 3.4.1

[larsoner]

This is the fix I also implemented elsewhere and it has worked well for the ~6 months I've used it, so in this goes. Thanks @rgommers !