Why do we need raw.execute?

[jordandsullivan]

If you just want to get an unmitigated expectation value for some relatively simple execute function, there appear to be two different equivalent ways of doing so:

raw.execute(circuit, executor, observable)
vs.
observable.expectation(circuit, executor)?

In the docstring for raw.execute, it states

mitiq/mitiq/raw/raw.py

Lines 21 to 24 in 42779ed

	The only purpose of this function is to provide the same interface for
	non-error-mitigated values as the rest of the techniques in Mitiq. This
	is useful when comparing error-mitigated results to non-error-mitigated
	results.

But it appears other methods do not have a pure execute function. They have execute_with_[qem_method] and mitigate_executor functions, but I don't see how having a raw.execute function is useful in maintaining parity across QEM methods.

It seems to me that it would be more logical to use observable.expectation(circuit, executor), since you can presumably define any executor you wish, with or without error mitigation. To me, this would maintain a more consistent interface.

[jordandsullivan]

I also think it's not immediately obvious from the name that raw.execute will give you an expectation value rather than, say, measurement counts.

[cosenal]

All the context seems to be in #1006

I sympathize with your comments though, there is basically three ways to do the same thing, and only for documentation purposes.

[jordandsullivan]

Yeah personally I'm not a fan of this much redundancy, since it makes the user waste time trying to figure out which API calls actually do something different vs. which ones are functionally an alias for another command.