Propagation of error for `pgd_process_tomo`?

[peterse]

If I have a matrix of POVM estimators and an identically-shaped matrix of systematic errors corresponding to each element (due to visibility, single-qubit gate infidelities for the state prep and readout gates, etc.), it would be cool if pgd_process_tomo would propagate error along with computing the reconstructed choi matrix.

Use case would be something like

reco_data, reco_err = pgd_process_tomo(M,A,info=true)
choi_reco = Operator(reco_data, (2,2,2,2))
choi_reco_err = Operator(reco_err , (2,2,2,2))

[jebej]

You would want the error on each element of the Choi matrix? I'm not sure how that would work. It seems tricky to calculate the derivative of the Choi matrix w.r.t. the input, given the optimization step in between.