Disallowing timedelta64+float promotion (to timedelta64) in all cases
(previously it was assymetric and "half allowed") meant that isclose,
allclose, np.ma.allclose, and assert_arrays_almost_equal (which uses
isclose), would stop work for timedelta64. Hardcoding that timedelta64
is passed on unmodified retains the old behaviour.

It may make sense to deprecate or change this behaviour in the future,
but for the 1.20 release, the behaviour should be as much unmodified
as possible.

Closes gh-18286

The `__new__` method of `MaskedArray` has the parameter `**options`,
but it is not used.  This means a mistyped parameter is silently ignored,
e.g. if I mistype the `mask` parameter like this, I don't get an error:

    In [2]: m = np.ma.MaskedArray([1, 2, 3], maks=[0, 1, 0])

    In [3]: m
    Out[3]:
    masked_array(data=[1, 2, 3],
                 mask=False,
           fill_value=999999)

In this PR, I simply remove `**options` from the signature of
`MaskedArray.__new__`.  This affected one test.  That test was passing the
argument `shape=(0, 1, 2)` to the constructor of a subclass of `MaskedArray`,
but since any unknown keywords were silently ignored, that use of `shape` had
no effect.  I replaced it with the positional argument `[[[]], [[]]]`, which
I believe has the effect that was originally intended with the use of the
`shape` argument.

* Conversion of some strings to fstrings

* Remove spaces

* Update numpy/ma/mrecords.py

Co-authored-by: Bas van Beek <43369155+BvB93@users.noreply.github.com>

* Update numpy/ma/tests/test_old_ma.py

Co-authored-by: Bas van Beek <43369155+BvB93@users.noreply.github.com>

* Update numpy/ma/timer_comparison.py

Co-authored-by: Bas van Beek <43369155+BvB93@users.noreply.github.com>

Co-authored-by: Jakob <jakobjakobson13@posteo.de>
Co-authored-by: Bas van Beek <43369155+BvB93@users.noreply.github.com>

* fix np.ma.core.doc_note

Fix for issue gh-15077. ma.core now checks and sets max and min
values for all float and complex dtypes. This fixes the max of a
masked array with a float16 or complex dtype returning inf

Builds on top of gh-15086.

Co-authored-by: William Jones <william.jones@physics.ox.ac.uk>

* BUG: fix AttributeError on accessing object in nested MaskedArray
Co-authored-by: Rena Xu <renaxu@Renas-MacBook-Pro.local>
Co-authored-by: Eric Wieser <wieser.eric@gmail.com>
Co-authored-by: vagrant <vagrant@vagrant.vm>

DEP: Deprecate ndarray.tostring()