[skip ci] Interpolate missing nominal values during Averaging

[sjperkins]

• Tests added / passed

  $ py.test -v -s africanus

  If the pep8 tests fail, the quickest way to correct
  this is to run autopep8 and then flake8 and
  pycodestyle to fix the remaining issues.

  $ pip install -U autopep8 flake8 pycodestyle
  $ autopep8 -r -i africanus
  $ flake8 africanus
  $ pycodestyle africanus
  

• Fully documented, including HISTORY.rst for all changes
  and one of the docs/*-api.rst files for new API

  To build the docs locally:

  pip install -r requirements.readthedocs.txt
  cd docs
  READTHEDOCS=True make html
  

[sjperkins]

@bennahugo would you consider giving this a quick review, mostly of the new test cases: test_interpolation and test_interpolation_edge? I'm less interested in review of the refactor, although you're welcome to provide feedback here too. Regardless I think the new test cases illustrate how the code now interpolates missing rows, as well as how it deals (imperfectly) with missing edge data.

Note that this doesn't change the BDA averaging code, yet. It only modifies the standard time and channel averaging code because it's far easier to reason about including X seconds worth of data in the bin. If you're happy with the standard case, I'll implement the same in the BDA code.

[sjperkins]

Discussion Points

• Regular/Irregular grid
• Missing/Non-missing data

10 seconds bin
3 seconds
X is missing (not even flagged)
O is present (flagged/unflagged)

O O O X X X X O

| O O O | X X X X O
| O O O | X X X | X O | -- split samples across multiple bins

| O O O | X X X X O | -- current strategy, samples completely in/out of a bin

| O O O | X X X X | O |