CLN: move away from numpy.fix in favor of numpy.trunc

[hellerve]

This PR replaces all occurrences of numpy.fix with numpy.trunc. This is in response to a currently discussed deprecation of numpy.fix as well as performance considerations (nice, but secondary).

This change should not change anything in a meaningful way, internally or externally.

It is, however, unclear to me whether this PR should also remove test_np_fix in test_ufunc.py. I opted to keep it. If desired, I can make this part of the PR. It is worth noting that no corresponding test_np_trunc exists (should it?).

pandas/pandas/tests/series/test_ufunc.py

Lines 460 to 467 in 3157d07

	@pytest.mark.xfail(reason="see https://github.com/pandas-dev/pandas/pull/51082")
	def test_np_fix():
	# np.fix is not a ufunc but is composed of several ufunc calls under the hood
	# with `out` and `where` keywords
	ser = pd.Series([-1.5, -0.5, 0.5, 1.5])
	result = np.fix(ser)
	expected = pd.Series([-1.0, -0.0, 0.0, 1.0])
	tm.assert_series_equal(result, expected)

Cheers

[mroeschke]

Thanks @hellerve. We can keep test_np_fix until numpy officially deprecates it

[hellerve]

Perfect. I’ll submit another PR once I get the deprecation PR at numpy through :)