TST: Remove dupe TimdeltaIndex tests from indexes/timedeltas/test_astype.py

[jschendel]

• tests added / passed
• passes git diff upstream/master -u -- "*.py" | flake8 --diff

Both classes in test_astype.py and test_timedelta.py inherit from DatetimeLike, which in turn inherits from Base, leading to 200+ dupe tests being run.

Modified test_astype.py to inherit from object instead, removed some unnecessary boilerplate, and removed some additional tests that were specified to override base class tests.

[jschendel]

note that this same test is also in test_arithmetic.py, and is overriden in test_timedeltas:

pandas/pandas/tests/indexes/timedeltas/test_timedelta.py

Lines 35 to 36 in 69cd5fb

	def test_shift(self):
	pass # this is handled in test_arithmetic.py

pandas/pandas/tests/indexes/timedeltas/test_arithmetic.py

Lines 170 to 187 in 69cd5fb

	def test_tdi_shift_int(self):
	# GH#8083
	trange = pd.to_timedelta(range(5), unit='d') + pd.offsets.Hour(1)
	result = trange.shift(1)
	expected = TimedeltaIndex(['1 days 01:00:00', '2 days 01:00:00',
	'3 days 01:00:00',
	'4 days 01:00:00', '5 days 01:00:00'],
	freq='D')
	tm.assert_index_equal(result, expected)
	def test_tdi_shift_nonstandard_freq(self):
	# GH#8083
	trange = pd.to_timedelta(range(5), unit='d') + pd.offsets.Hour(1)
	result = trange.shift(3, freq='2D 1s')
	expected = TimedeltaIndex(['6 days 01:00:03', '7 days 01:00:03',
	'8 days 01:00:03', '9 days 01:00:03',
	'10 days 01:00:03'], freq='D')
	tm.assert_index_equal(result, expected)

[codecov]

Codecov Report

Merging #19509 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master   #19509   +/-   ##
=======================================
  Coverage   91.69%   91.69%           
=======================================
  Files         148      148           
  Lines       48561    48561           
=======================================
  Hits        44530    44530           
  Misses       4031     4031

Flag	Coverage Δ
#multiple	90.06% <ø> (ø)	⬆️
#single	41.72% <ø> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 69cd5fb...1a43bb5. Read the comment docs.

[jbrockmendel]

+1

FWIW there are some other astype tests that may be in suboptimal locations:

test_timedelta.test_freq_conversion (and the Series cases probably belong elsewhere)
test_ops.test_astype_object

[jreback]

remove any of these _multi precess_can_split, they are really old

[jreback]

actually let's do that separately.

[jreback]

thanks!