Add `initial_state` to `_post_circuit`

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (463aa85) 100.00% compared to head (c15c2df) 100.00%. Report is 7 commits behind head on main.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #27 +/- ## ========================================= Coverage 100.00% 100.00% ========================================= Files 4 4 Lines 119 120 +1 ========================================= + Hits 119 120 +1 ``` | [Flag](https://app.codecov.io/gh/qiboteam/qibo-client/pull/27/flags?src=pr&el=flags&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=qiboteam) | Coverage Δ | | |---|---|---| | [unittests](https://app.codecov.io/gh/qiboteam/qibo-client/pull/27/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=qiboteam) | `100.00% <100.00%> (ø)` | | Flags with carried forward coverage won't be shown. [Click here](https://docs.codecov.io/docs/carryforward-flags?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=qiboteam#carryforward-flags-in-the-pull-request-comment) to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[scarrazza]

The initial state is the first argument when executing a circuit (before the nshots option).

[marcorossi5]

I see. How do you want to pass it to Django? JSON? If np.ndarray is an option you might want to serialize to bytes with pickle. My question is: is it safe to execute pickled objects received from http requests?

[scarrazza]

That is a good point, @MatteoRobbiati does ibmq supports this feature? Usually passing initial states are not possible for hardware execution. If this is the case we should not support this feature.

[alecandido]

If np.ndarray is an option you might want to serialize to bytes with pickle.

There is a safe (and portable) way to serialize NumPy arrays to bytes, i.e. np.save.

Unfortunately, it is writing to a file-like object, so you have to do a little bit of gymnastics to get it in memory (without passing through disk), but it's not terrible. It's something like:

import io

buffer = io.BytesIO()
np.save(buffer, myarray)
buffer.seek(0)
mybytes = buffer.read()

and io is standard library.

I didn't check it, but if that doesn't work should be something very close.

[alecandido]

Usually passing initial states are not possible for hardware execution.

For hardware, the only way would be to disassemble it as a circuit, isn't it?

[scarrazza]

Indeed, so given that the major goal of this project is to execute on quantum hardware, we should give up on this setup which might be even more complex, i.e. if a crazy user decides to send 1TB of state vector...

[MatteoRobbiati]

I agree with the direction you suggested. Together with @BrunoLiegiBastonLiegi we tested the Qiskit backend and it seems they adopt the "hardware-like" approach (circuits as input, frequencies as output). Fine with closing this and, then, we should just remove the initial_state argument from the qibo-cloud-backend/qibo_client.py.

[scarrazza]

Thanks @MatteoRobbiati for checking, please go ahead with changes in the qibo-cloud-backends.