Performance Improvement

[leilabbb]

Verified suggestions by copilot. The main ones are:

• added " if statements " to catch extreme cases
• fixed function return to formulate lists

[leilabbb]

pre-commit.ci autofix

[benjwadams]

The removed version will never be true since numpy.unique returns a 1D array if the axis argument is left unspecified or None. https://numpy.org/doc/stable/reference/generated/numpy.unique.html

The second added version (which admittedly Copilot suggested) will fail and throw an exception for anything but a single-element array, which is not going to be the majority of datasets/variables. I would probably just compare unique_vals == np.array([inp._FillValue]) or something similar, but I don't know how this would operate for masked arrays.

>>> import numpy as np
>>>arr = np.array([1, 2])
>>> arr.item()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: can only convert an array of size 1 to a Python scalar

Please address and add unit tests for this case.

[benjwadams]

Commit message should be more specific. Right now I wouldn't know what code the changes affected looking at the commit summary.

- added " if statements " to catch extreme cases
- fixed function return  to formulate lists

I would change this to something like:

- Added conditional statements to only remove QARTOD variables if thresholds are unset
- Append to list of diagnostic messages returned by QC routines when masked data is detected

Amending the commit message and force pushing would be OK.