pyarrow: Support date32[day] and date64[ms] dtypes in pandas objects

[weiji14]

Description of proposed changes

Handle date columns in pandas.DataFrame with pyarrow dtypes like date32[day][pyarrow] or date64[ms][pyarrow].

This is implemented by modifying the vectors_to_arrays conversion function to convert date32/date64 dtypes to numpy.datetime64.

An altenative implementation would be to handle pyarrow date32/date64 dtypes natively without converting, but this might require more work and discussion at #2800.

Note:

• Unlike numpy which only has datetime64, PyArrow has separate Date types (date64/date32) and Time types (time64).

References:

• https://arrow.apache.org/docs/python/pandas.html#date-types
• https://arrow.apache.org/docs/13.0/python/generated/pyarrow.date32.html
• https://arrow.apache.org/docs/13.0/python/generated/pyarrow.date64.html

Part of #2800, related to #242, extends #464

Reminders

• Run make format and make check to make sure the code follows the style guide.
• Add tests for new features or tests that would have caught the bug that you're fixing.
• Add new public functions/methods/classes to doc/api/index.rst.
• Write detailed docstrings for all functions/methods.
• If wrapping a new module, open a 'Wrap new GMT module' issue and submit reasonably-sized PRs.
• If adding new functionality, add an example to docstrings or tutorials.
• Use underscores (not hyphens) in names of Python files and directories.

Slash Commands

You can write slash commands (/command) in the first line of a comment to perform
specific operations. Supported slash commands are:

• /format: automatically format and lint the code
• /test-gmt-dev: run full tests on the latest GMT development version

[seisman]

Looks good to me, but need to mark the PR as ready for review and merge the main branch to make sure all tests pass.

[weiji14]

I'm still debating on whether to add pyarrow into test_plot_datetime that was added in #464:

pygmt/pygmt/tests/test_plot.py

Lines 409 to 452 in 20054a1

	@pytest.mark.mpl_image_compare
	def test_plot_datetime():
	"""
	Test various datetime input data.
	"""
	fig = Figure()
	fig.basemap(
	projection="X15c/5c",
	region=[
	np.array("2010-01-01T00:00:00", dtype=np.datetime64),
	pd.Timestamp("2020-01-01"),
	0,
	10,
	],
	frame=True,
	)
	# numpy.datetime64 types
	x = np.array(
	["2010-06-01", "2011-06-01T12", "2012-01-01T12:34:56"], dtype="datetime64"
	)
	y = [1.0, 2.0, 3.0]
	fig.plot(x=x, y=y, style="c0.2c", pen="1p")
	# pandas.DatetimeIndex
	x = pd.date_range("2013", freq="YS", periods=3)
	y = [4, 5, 6]
	fig.plot(x=x, y=y, style="t0.2c", pen="1p")
	# xarray.DataArray
	x = xr.DataArray(data=pd.date_range(start="2015-03", freq="QS", periods=3))
	y = [7.5, 6, 4.5]
	fig.plot(x=x, y=y, style="s0.2c", pen="1p")
	# raw datetime strings
	x = ["2016-02-01", "2017-03-04T00:00"]
	y = [7, 8]
	fig.plot(x=x, y=y, style="a0.2c", pen="1p")
	# the Python built-in datetime and date
	x = [datetime.date(2018, 1, 1), datetime.datetime(2019, 1, 1)]
	y = [8.5, 9.5]
	fig.plot(x=x, y=y, style="i0.2c", pen="1p")
	return fig

That test is already a bit long though ...

[weiji14]

I'm still debating on whether to add pyarrow into test_plot_datetime that was added in #464:

On second thought, I might do this in a follow-up PR. PyArrow also has time32/time64 types, so I might add a test_plot_datetime_pyarrow unit test that reuses test_plot_datetime's baseline image, but using pyarrow inputs instead.