Open braised-babbage opened 5 years ago
Are the tests pulling data from https://forest-server.qcs.rigetti.com?
It looks like if a gate fidelity is not present, then it is assigned PERFECT_FIDELITY = 1.0
https://github.com/rigetti/pyquil/blob/master/pyquil/device.py#L565
Ah, so this is a quilc
thing...
(defconstant +near-perfect-fidelity+ 0.99999d0
"Even perfect operations are typically limited in their physical realization by, say, the granularity of control electronics. (For instance, waveform IQ values might be stored as complex fixnums of some specified depth.) This constant is a mnemonic for \"supposedly perfect\" and captures some of the loss incurred by these imperfections.")
Ah, so this is a
quilc
thing...(defconstant +near-perfect-fidelity+ 0.99999d0 "Even perfect operations are typically limited in their physical realization by, say, the granularity of control electronics. (For instance, waveform IQ values might be stored as complex fixnums of some specified depth.) This constant is a mnemonic for \"supposedly perfect\" and captures some of the loss incurred by these imperfections.")
As I understand it (almost verbatim from @ecpeterson), this is because the compiler will for
Issue Description
Several tests get this warning. It is not apparent why, considering the
specs_dict
inconftest.py
does not contain any fidelities greater than 1.How to Reproduce
To pick one example: