Adds target kwarg to tomography init #1025
Conversation
This allows specifying a custom target for analysis when initializing the experiments. If set to None automatic target calculation is disabled.
There was a problem hiding this comment.
Nice, looks good.
The one thing you could add is a test of passing a non-None target. That would mostly be to protect against a future code change breaking the feature since it is pretty clear it would work fine with the current code.
One question about #994 -- I think the original problem was the user passing a circuit already mapped to high qubit number. Is there any way to invoke the simulator so that it ignores the idle qubits and could simulate qubits 16 and 17 without using 2**17 by 2**17 matrices?
Co-authored-by: Will Shanks <wshaos@posteo.net>
|
@wshanks Its possible, but more work so I skipped it for now. When permuting the circuit based on the measurement and preparation qubits you could trim any uncoupled ancilla qubits before simulating the statevector/unitary, rather than simulate them all and trace them out at the end. If I get time before release I might try and implement that. |
This allows specifying a custom target for analysis when initializing the experiments. If set to None automatic target calculation is disabled. Co-authored-by: Will Shanks <wshaos@posteo.net>
Summary
Closes #994
Updates State and process tomography experiments to directly set the target option for TomographyAnalysis rather than pass it through experiment data metadata.
Details and comments