Enhanced ValueError messages

[romainfd]

When using values out of range (examples below for qubit positions and pulse detuning), the ValueError exception could indicate the minimum/maximum authorized value to make fixing the issue faster. The value is used in the test just above so printing it looks easy at first glance.

• Illustration for min qubits distance
  

• Illustration for max pulse detuning
  

[HGSilveri]

Hi @romainfd , thanks for the suggestion! I will tag it as a "good first issue" for whoever wants to tackle it.

[LaurentAjdnik]

Hi! I've been doing so for the new functions Register.hexagon() and Register.max_connectivity() (soon to be PR-ed), so I might just as well continue for others.

Fortunately, some of them already do, like Channel.validate_duration() for instance.

Anyhow, this will be a good opportunity to:

• Make sure allowed values are always mentioned as part of Error messages
• Harmonize the way they are included in strings (throughout the code, you'll find F-strings, .format() and perhaps others)

By the way, I found almost no checks for None on parameters. Testing against allowed values will result in a TypeError (see Ordering comparisons).

Do we want to add checks and throw an explicit ValueError with a message rather than this default behavior ? This is already the case in Register.draw(), SimulationResults.sample_final_state() and Variable.build() but nowhere else.

[HGSilveri]

@LaurentAjdnik thanks! I agree with your suggestions, although I think it would be better to do them in a separate PR if possible, otherwise the one you're currently working on might get too big.

Regarding checks for None on parameters, those do sound a bit of overkill to me, but I will accept them if you feel they are necessary.

[LaurentAjdnik]

@HGSilveri: You can assign this one to me, as it is related to #157.

[HGSilveri]

@LaurentAjdnik If you're satisfied with the changes from your last PR, I'll close this, just let me know.

[LaurentAjdnik]

@LaurentAjdnik If you're satisfied with the changes from your last PR, I'll close this, just let me know.

Yup, let's close this one. We might PR a few adjustments in the future but no need to keep this issue open.