Add test for multi-antennas

[ToddSchumannNTIA]

No actual code changes needed to implement multi-antenna paths for the differential calibration. Just added a pytest for the new format.

Testing with two actions with a dummy differential calibration file with 10dB difference between the two antenna paths:

scos-sensor-api-1  | [2024-07-17 15:54:39,193] [DEBUG] Using sensor gain: 8.75119720678694 dB
scos-sensor-api-1  | [2024-07-17 15:54:39,193] [DEBUG] Using differential loss: 10.318735684035396 dB
scos-sensor-api-1  | [2024-07-17 15:54:39,193] [DEBUG] Applying total gain of -1.5675384772484549

scos-sensor-api-1  | [2024-07-17 15:56:15,601] [DEBUG] Using sensor gain: 30.097933572513824 dB
scos-sensor-api-1  | [2024-07-17 15:56:15,601] [DEBUG] Using differential loss: 20.538227261071448 dB
scos-sensor-api-1  | [2024-07-17 15:56:15,601] [DEBUG] Applying total gain of 9.559706311442376

[aromanielloNTIA]

I'm not sure why my last commit didn't trigger Tox running on this branch, but I manually ran it on that commit. Tests all passed. Results available here.

The issue was unrelated to the changes in this PR- some SciPy functions changed to support fewer positional arguments, favoring keyword arguments, in v1.12.0. This caused the Tox testing with newer Python versions to fail. I updated every SciPy function to use only keyword arguments since it seems they might make similar changes in the future (e.g. here and here). This should help future-proof us from having similar issues down the line.

I also updated pre-commit hooks to latest versions and ran on all files.

[aromanielloNTIA]

Looks good- thanks for adding the test case for this feature which was already supported but not tested. Testing on sensor (seadog02) also verified the changes.