API: properly deprecate changes in to_data_frame

[drammock]

This PR adds a proper deprecation cycle to the API changes in #7206

It also restores parameter order to more closely match what it used to be, to minimize downstream code breakage with positional argument passing.

[larsoner]

LGTM +1 for merge

[codecov]

Codecov Report

❗ No coverage uploaded for pull request base (master@9f3b76f). Click here to learn what that means.
The diff coverage is 100%.

@@            Coverage Diff            @@
##             master    #7326   +/-   ##
=========================================
  Coverage          ?   89.88%           
=========================================
  Files             ?      450           
  Lines             ?    81251           
  Branches          ?    12915           
=========================================
  Hits              ?    73033           
  Misses            ?     5396           
  Partials          ?     2822

[drammock]

Travis failures are unrelated:

Xdawn:

       3 failed
         - mne/inverse_sparse/tests/test_mxne_inverse.py:34 test_mxne_inverse_standard
         - mne/preprocessing/tests/test_xdawn.py:63 test_xdawn_fit
         - mne/utils/tests/test_numerics.py:239 test_reg_pinv

mne/preprocessing/xdawn.py:189: in _fit_xdawn

    evals, evecs = linalg.eigh(evo_cov, signal_cov)

../../../virtualenv/python3.7.1/lib/python3.7/site-packages/scipy/linalg/decomp.py:491: in eigh

    " computed." % (info-b1.shape[0]))

E   numpy.linalg.LinAlgError: the leading minor of order 8 of 'b' is not positive definite. The factorization of 'b' could not be completed and no eigenvalues or eigenvectors were computed.

Numerics:

mne/utils/tests/test_numerics.py:270: in test_reg_pinv

    assert_array_equal(a_inv_np, a_inv_mne)

E   AssertionError: 

E   Arrays are not equal

E   

E   Mismatched elements: 1 / 9 (11.1%)

E   Max absolute difference: 2.77555756e-17

E   Max relative difference: 2.22044605e-16

E    x: array([[ 0.375   ,  0.      , -0.125   ],

E          [ 0.      ,  0.333333,  0.      ],

E          [-0.125   ,  0.      ,  0.375   ]])

E    y: array([[ 0.375   ,  0.      , -0.125   ],

E          [ 0.      ,  0.333333,  0.      ],

E          [-0.125   ,  0.      ,  0.375   ]])

MXNE:

mne/inverse_sparse/tests/test_mxne_inverse.py:154: in test_mxne_inverse_standard

    pick_ori='vector')

</home/travis/build/mne-tools/mne-python/mne/externals/decorator.py:decorator-gen-436>:2: in tf_mixed_norm

    ???

mne/utils/_logging.py:90: in wrapper

    return function(*args, **kwargs)

mne/inverse_sparse/mxne_inverse.py:630: in tf_mixed_norm

    weights, weights_min)

mne/inverse_sparse/mxne_inverse.py:69: in _prepare_gain

    use_cps=True, **depth)

mne/minimum_norm/inverse.py:1387: in _prepare_forward

    gain = np.dot(whitener, forward['sol']['data'])

<__array_function__ internals>:6: in dot

    ???

E   Failed: Timeout >120.0s

[agramfort]

error is unrelated. Maybe due to some update of numpy/scipy on conda a few days ago

https://anaconda.org/anaconda/scipy/files
https://anaconda.org/anaconda/numpy/files

[larsoner]

error is unrelated. Maybe due to some update of numpy/scipy on conda a few days ago

FYI I have some of these fixed in #7325, I'll try to squash the last failure or two there