Should be able to do measurement mitigation on multiple independent basis sets at same time

[nonhermitian]

What is the expected behavior?

The measurement mitigation functions should be able to simultaneously correct errors over multiple independent qubit sets on a device using a single job (if possible). For example, it should be possible to simultaneously mitigate measurement error over all pairs of qubits in the coupling map of a device.

[dcmckayibm]

It's been a while since I tested it, but it should be able to do that. When you define the tensored mitigation you can define sets of qubits. The problem is that the correction still scales badly (and it's not a trivial problem to fix).

Also defining based on a coupling map doesn't work because there are no independent pairs (they are all coupled).

[nonhermitian]

I should say I want to do all pairs defined by the coupling map in a single job and get out a different fitter/filter for each pair

[dcmckayibm]

Let's say the coupling map is [[0,1],[1,2]] you can currently calibrate a fitter for [[0,1],[2]] (fitter1) and [[0],[1,2]] (fitter2) I just don't think you'll find it's much more useful than just tensoring all the qubits independentally. Because you either apply Fitter1 or Fitter2 and I'm not sure how to choose.