Int Index.slice_locs Not Raising with Float Arguments

[WillAyd]

xref #4892 as I'm going through #20812 I am seeing tests which suggest that trying to call slice_locs with float arguments against an integer type Index should raise a TypeError:

pandas/pandas/tests/indexes/test_base.py

Line 1350 in c4da79b

pytest.raises(TypeError,

While these tests are currently passing, I believe the TypeError is getting raised from the call to pytest.raises (specifically a 'list' object is not callable for the third argument to pytest.raises) and not as a result of the actual index.slice_locs call.

>>> pd.__version__
'0.23.0.dev0+824.gc4da79b5b.dirty'
>>> index = pd.Index(np.array([0, 1, 2, 5, 6, 7, 9, 10], dtype=int))
>>> index.slice_locs(5.0, 10.0)
(3, 8)
>>> index.slice_locs(4.5, 10.5)
(3, 8)

[simonjayhawkins]

the TypeError message was added in #4850

https://github.com/jreback/pandas/blob/60efe85bc2ce9578e8f5e0c88907c5cc6b6b7b31/pandas/core/index.py#L526-L531

and has since been removed. int slicing with floats is now allowed.