[FIX] remove indexing along channel axis in AnalogSignalGap.load()

[KristijanArmeni]

Reference issue

Fixes #12356.

What does this implement/fix?

Removing indexing along channel dimension with idx in AnalogSignalGap.load() on zero-arrays representing the gap data in self.signal (commit: 1aa66). These arrays are constructed on the fly using the idx variable and hence have the appropriate channel dimension to begin with. In case of a single channel, slicing a 1-dimensional array with slice(1:2) slice yields a 0-dimensional array which breaks the concatenation downstream in l. 359.

[larsoner]

pip-pre will fail until pyvista/pyvista#5460 lands so you can ignore it

[larsoner]

It would be great to have a test that would have caught this, something to keep in mind if you ever come across a suitably small dataset @KristijanArmeni !

[KristijanArmeni]

Thanks, @larsoner!

Can write a test at this point -- we already have 2 Neuralynx channels that simulate gaps, so I can reproduce this bug on the testing dataset.

The code below fails on current main (but will pass on this branch):

(3 durations < 0.005s hidden.  Use -vv to show these durations.)
===================================== short test summary info ======================================
FAILED mne/io/neuralynx/tests/test_neuralynx.py::test_neuralynx_gaps - Failed: Failed to load data after picking one channel with gaps: all the input array dimensions...

# test that channel selection works
raw = read_raw_neuralynx(
    fname=testing_path,
    preload=False,
    exclude_fname_patterns=ignored_ncs_files,
)

raw.pick("LAHC2")
try:
    raw.load_data()
except Exception as e:
    pytest.fail(f"Failed to load data after picking one channel with gaps: {e}")

[larsoner]

Can write a test at this point -- we already have 2 Neuralynx channels that simulate gaps, so I can reproduce this bug on the testing dataset.

Great, feel free to push a version of that here then!

[KristijanArmeni]

Great, feel free to push a version of that here then!

Ok, added test in edaee01. @larsoner

[larsoner]

Marking for merge when green, thanks in advance @KristijanArmeni !