Missing Unit Tests

[purva-thakre]

Noticed some lines in the following files show up as uncovered by unit tests in the pytest coverage report. This issue is to figure out if these lines require additional unit tests or if they erroneously show up as uncovered in the pytest report.

• mitiq/_about.py: L 21-59, 63
• mitiq/benchmarks/mirror_circuits.py: L 155
• mitiq/benchmarks/randomized_benchmarking.py: L 54
• mitiq/benchmarks/randomized_clifford_t_circuit.py: L 43, 47
• mitiq/calibration/calibrator.py: L 385, 432
• mitiq/calibration/settings.py: L 185, 244-246, 269, 318
• mitiq/cdr/_testing.py: L 11-15
• mitiq/cdr/clifford_training_data.py: L 232
• mitiq/cdr/clifford_utils.py: L 79-81
• mitiq/executor/executor.py : L 141, 205
• mitiq/interface/mitiq_braket/conversions.py : L 324
• mitiq/interface/mitiq_qiskit/conversions.py : L 169
• mitiq/observable/observable.py: L 84, 89
• mitiq/observable/pauli.py: L 48, 55, 122, 166, 178
• mitiq/pec/representations/optimal.py: L 149
• mitiq/pec/types/types.py : L 239
• mitiq/shadows/quantum_processing.py: L 140 Ignored temporarily due to Make changes to the functions in mitiq.shadows #2129
• mitiq/shadows/shadows.py: L 68, 73, 78, 137 Ignored temporarily due to Make changes to the functions in mitiq.shadows #2129
• mitiq/shadows/shadows_utils.py: L 61, 88, 91-96 Ignored temporarily due to Make changes to the functions in mitiq.shadows #2129
• mitiq/zne/inference.py: L 990, 1253, 1331, 1333, 1335, 1415, 1454, 1460-1461, 1588
• mitiq/zne/scaling/folding.py: L 119, 294
• mitiq/zne/zne.py: L 59

Will update the issue description later with more details about the specific lines.

[cosenal]

Happy to go through this exercise, with the caveat that one could make the The Case Against 100% Code Coverage.

[purva-thakre]

The caveat is understandable. I don't go by the coverage percentage. That value fluctuates inconsistently and it's still not clear to me how it is calculated.

Typically, I look at the uncovered lines in the pytest coverage report to make pytest ignore some lines from the coverage report or add unit tests. Perhaps we should discuss what type of unit tests should be added and which ones could be ignored.

For example, I don't think we need unit tests for mitiq/_about.py but # pragma: no cover is ignored by the pytest coverage report. So, we will have to create a PR to make pyest ignore these.

mitiq/mitiq/_about.py

Lines 26 to 29 in aefc5ce

	try:
	from qiskit import __qiskit_version__ # pragma: no cover
	qiskit_version = __qiskit_version__["qiskit"] # pragma: no cover

[cosenal]

That value fluctuates inconsistently

It should no longer fluctuate, after #2319, which fixed #2318

[purva-thakre]

Closely related to this issue, the following lines are ignored in the pytest coverage report. A unit test could be added to raise the error similar to some of the changes in #2366

Unless I am missing something and we can't raise this error?

mitiq/mitiq/pec/representations/damping.py

Lines 75 to 77 in aefc5ce

	raise ValueError( # pragma: no cover
	"Only single-qubit operations are supported." # pragma: no cover
	) # pragma: no cover