Open caleb-johnson opened 8 months ago
It would be nice to leave our QPDBasis decomps specified as-is, as they tend to follow the literature more directly.
We may allow users to specify a basis gate set for which to specify decompositions. The default can be the eagle/heron basis set. When ckt generates experiments, it will translate the QPDBasis decomps into the specified basis set before adding them to the subexperiment circuit
The equivalence library in Qiskit may give us a nice interface through which we can translate our gates before applying them to the circuit in generate_cutting_experiments
, partition_problem
, etc.
generate_cutting_experiments
, partition_problem
, etc. may take an optional basis gate set as input
If our cutting decompositions were specified in terms of Eagle/Heron basis gates, it would give users the opportunity to transpile one circuit and then run all the subexperiments without having to transpile so many circuits