Gauge optimization seems to ignore instruments

[rileyjmurray]

Corey left the following comment on PR#410:

I think there might be a (likely preexisting) bug here related to instruments. For the fidelity and tracedistance metrics I’m not seeing the instrument elements being accessed (except in passing in the spam metric computation where it is just used to filter the item weights). For frobenius this is harder to see, but I took a look at the implementation of the ‘frobeniusdist’ method for ExplicitOpModels here and I’m not seeing instruments there either.

Originally posted by @coreyostrove in #410 (comment)
See also, #410 (comment).

We should revise gauge optimization code so loss functions include contributions from instruments.

[rileyjmurray]

Good news! Gauge optimization with respect to (squared) Frobenius distance already considers instruments.

This is because gauge optimization w.r.t. Frobenius distance is mediated through either ExplicitOpModelCalc.frobeniusdist or ExplicitOpModelCalc.residuals:

and ExplicitOpModelCalc objects are constructed in a way that lumps instruments with gates

Gauge optimizing w.r.t. subspace-restricted Frobenius distance only uses ExplicitOpModelCalc.frobeniusdist, and incorporates the subspace information in the "transform_mx" arg.

Gauge optimization with respect to trace distance or fidelity does not currently consider instruments. I'll open a PR with fixes for these.